0.17 pycon-lithuania-2025 2025-04-22 2025-04-26 5 00:05 https://pretalx.com Europe/Vilnius 101 Talk 2025-04-23T09:00:00+03:00 09:00 00:25 A pycon-lithuania-2025-68682-python-day-opening Org en false https://pretalx.com/pycon-lithuania-2025/talk/D39BNT/ https://pretalx.com/pycon-lithuania-2025/talk/D39BNT/feedback/ 101 Keynote 2025-04-23T09:30:00+03:00 09:30 01:00 The talk has two parts, show and tell what happened or happening in our world by the use of technology, the same technology we, the people in this room create. And also to tell you all that together we have the power to change this. **We can show solidarity to other human beings and groups** and fight back to evil and morally wrong technology decisions. pycon-lithuania-2025-68221-ethics-privacy-and-few-other-words Keynote Kushal Das en The talks goes through roughly the following examples: - IBM and holocaust - Data and people destroying data to save others during second world war - metadata - metadata based attacks in modern days - enshittification - social media based troubles, effects in elections - fight against e2e encryption - techbros trying to define what is a family - amazon and voice recordings - cryptocurrencies - AI and many examples of that land - surveillance industry - russia changing internet access in occupied Ukraine - china - climate effect - Examples of doing things in a good way, technologists fighting back, showing us hope false https://pretalx.com/pycon-lithuania-2025/talk/FRPUZS/ https://pretalx.com/pycon-lithuania-2025/talk/FRPUZS/feedback/ 101 Talk 2025-04-23T11:00:00+03:00 11:00 00:25 Code review is central part of everyday developer job. The motivation to create this talk was a quote: “The most important superpower of a developer is complaining about everybody else's code”. In this talk I’ll explain my approach to better code review. pycon-lithuania-2025-62625-code-review-the-right-way Python Day - Apr 23 Andrii Soldatenko en Sometimes it’s hard to convince a colleague about change or don’t change some lines of code, in my talk I would like to cover some best practices from my software engineering experience about efficient and honest code review. How to create culture of perfect code review. How to apply automatic tools to improve code review routine of repetitive comments or suggestions. How to write/or reuse codign style guides for you team to reduce time of arguing about naming conventions and different styles. What’s need to be automated and what’s need to be not automated during code review. The key role of patterns which can be reusable to not confuse colleagues. false https://pretalx.com/pycon-lithuania-2025/talk/98VE9C/ https://pretalx.com/pycon-lithuania-2025/talk/98VE9C/feedback/ 101 Talk 2025-04-23T11:30:00+03:00 11:30 00:25 Struggling with slow I/O in your Django apps? Want to maximize server resources? This talk explores asynchronous Python and its impact on Django. Let's clarify parallel vs. concurrent programming, and demystify Python's concurrency model, focusing on coroutines and the event loop. Learn how asyncio enables efficient, non-blocking code, handling concurrent requests without thread/process overhead and how everything is integrated into the Django framework. pycon-lithuania-2025-64231-sync-or-async-feel-the-magic-of-coroutines-and-the-event-loop-in-django Python Day - Apr 23 Antonis Kalipetis en Are you struggling with slow I/O operations in your Django applications? Do you want to unlock the full potential of your server's resources? This talk dives into the world of asynchronous programming in Python and explores how it can work for your Django applications! Let's clarify parallel vs. concurrent programming, and demystify Python's concurrency model, focusing on coroutines and the event loop. Learn how asyncio enables efficient, non-blocking code, handling concurrent requests without thread/process overhead and how everything is integrated into the Django framework. The core of the talk will focus on practical applications within Django. We'll explore what parts of a Django application can benefit most from asynchronous execution, including handling external API calls, database interactions, and long-running tasks. We'll demonstrate how to integrate asyncio into your Django views and other components, showcasing real-world examples and best practices. Finally, we'll discuss the benefits of running Django in asynchronous mode. We'll cover performance improvements, increased throughput, and better resource utilization, demonstrating how asynchronous Django can lead to more responsive and scalable web applications. Join this talk to discover the magic of coroutines and the event loop and learn how to bring the power of asynchronous programming to your Django applications. false https://pretalx.com/pycon-lithuania-2025/talk/BXDLW7/ https://pretalx.com/pycon-lithuania-2025/talk/BXDLW7/feedback/ 101 Talk 2025-04-23T12:00:00+03:00 12:00 00:25 "Trusted publishing" is the term for using the OpenID Connect (OIDC) standard in the Python Ecosystem to release on PyPI. In this talk will go though the usage of trusted publishing in any Python project and how it helped Ansible project to open up release management to the community. This talk is a deep dive explanation of release practicalities of releasing Ansible using trusted publishing. pycon-lithuania-2025-66095-using-trusted-publishing-to-ansible-release Python Day - Apr 23 /media/pycon-lithuania-2025/submissions/EDKURL/anwesha2_BC0t8bK.jpg Anwesha Das en "Trusted publishing" is the the way of exchanging short-lived identity tokens between a trusted third-party service and PyPI. This key feature in PyPI empowers the project maintainers to make releases via automated environments directly . This helps us to get rid of the use of manually generated API tokens. This talk will dig deeper in the practical aspects and impact of moving manual release process to automated release via github actions and trusted publishing. The talk will describe the trusted publishing from the view of a Release Manager of a critical project like Ansible. In my talk, I will go through. What is trusted publishing? why it is needed? How to use trusted publishing? Ansible manual release process in nutshell. Releasing Ansible with Github actions and Trusted Publishing Release automation: lessons learned true https://pretalx.com/pycon-lithuania-2025/talk/EDKURL/ https://pretalx.com/pycon-lithuania-2025/talk/EDKURL/feedback/ 101 Talk 2025-04-23T14:00:00+03:00 14:00 00:25 If you work with web services, you’re probably using containers… and you’re also probably not doing it as well as you could. In this talk, we’ll go over best practices for container images to produce lightweight, safe and modular containers for quick and efficient builds. pycon-lithuania-2025-61738-python-containers-best-practices Python Day - Apr 23 /media/pycon-lithuania-2025/submissions/3AZERE/containers_cAlJvv1.jpg Daniel Hervás en Containers are ubiquitous in today’s world. Almost everyone uses them, but how many of them are as well built as they should? Images usually border the line between backend and ops work, which means that the expertise might be diluted. Added to the fact that the symptoms of suboptimal images can be hard to spot at first glance, this point is usually a blind spot for many engineering teams. On the flip side, the benefits of having slimmer images isn’t just size for size’s sake, it’s smaller surface area for vulnerabilities, quicker builds which mean lower cost and ramp-up time, and faster re-builds in development or testing environments leading to quicker development iteration. In this talk, we’re gonna take a look at all the different ways we can optimize our Python image builds. Differences between base images, image composition, demystifying multi-stage builds, removing unnecessary packages, dependencies and caches from the final result, the build cache and how to best use it, etc. We’ll see all of this and more, and finally, we’ll wrap up with quick overviews of extra tooling that can be used for extremely small sizes, and notes for multi-arch builds. false https://pretalx.com/pycon-lithuania-2025/talk/3AZERE/ https://pretalx.com/pycon-lithuania-2025/talk/3AZERE/feedback/ 101 Talk 2025-04-23T14:30:00+03:00 14:30 00:25 Programming is for a large about removing repetition and finding abstractions that achieve that. But is that always sensible? Using the power of music we will examine how universal DRY really is. pycon-lithuania-2025-57176-do-repeat-yourself Python Day - Apr 23 /media/pycon-lithuania-2025/submissions/QW3LGC/banner_EPdD4ug.png Donatas Rasiukevičius en Don't Repeat Yourself (or DRY) is common advice for programmers. While this can be useful in some cases, following this dogmatically can put you in a bad spot. This talk dives into the painful lessons that can arise from this approach, how it can complicate code and increase cognitive load. Beyond that, we also examine how repetition can facilitate both clarity and learnability and how it can be less evil than it is sometimes made out to be. We'll examine all this through the lens of music programming in Python. false https://pretalx.com/pycon-lithuania-2025/talk/QW3LGC/ https://pretalx.com/pycon-lithuania-2025/talk/QW3LGC/feedback/ 101 Talk 2025-04-23T15:00:00+03:00 15:00 00:25 Coding aesthetics, in this context, refers to how code is written. It is essential that programmers also pay attention to the aesthetics and not just the functionality the code aims to achieve. This talk explores several ways to make Python code aesthetically pleasing, such as code refactoring, using static code analysis tools like PyLint to check compliance with PEP8 guidelines, and applying syntactic sugar. In addition, we will discuss the limitations of PEP8 and how we can make more pragmatic choices. pycon-lithuania-2025-64658-coding-aesthetics-pep-8-existing-conventions-and-beyond Python Day - Apr 23 Shiva Bhusal en The talk will start with a few questions, and also with Donald Knuth's quote: "Programming is the art of telling another human being what one wants the computer to do." It will comprise of 4 sections: A. Intro (1 mins) B. Donald Knuth's quote and what coding aesthetics means in this context (4 mins) C. Ways to maintain coding aesthetics (10 mins in total) - Refactoring: 3 mins - PEP 8 and Code Analysis Tools ( PyLint, Flake8, Black, etc.): 4 mins - Syntactic Sugars in Python: 3 mins D. Beyond PEP8: 5 mins - Purity vs Pragmatism - Examples where PEP8 could be limiting - Beyond PEP8 E. Q&A: 5 mins false https://pretalx.com/pycon-lithuania-2025/talk/7QWCRS/ https://pretalx.com/pycon-lithuania-2025/talk/7QWCRS/feedback/ 101 Keynote 2025-04-23T16:00:00+03:00 16:00 01:00 The modern Python programmer spends little time thinking about the classic ‘Design Patterns’ from the 1990s. Why are they no longer relevant? This keynote address will explore how we write Python code today, and how it avoids the problems that design patterns were meant to solve. pycon-lithuania-2025-66935-skip-the-design-patterns-architecting-with-nouns-and-verbs Keynote /media/pycon-lithuania-2025/submissions/3RPFYL/proposal_CyGPhFY.png Brandon Rhodes en Why does the old 1990s _Design Patterns_ book still appear on lists of books for programmers to read, when so many practicing coders are able to work for years at a time without even thinking about them? Isn’t it time for the book to be stricken from the list? Several years ago, to prevent modern Python programmers from accidentally using old and out-of-date patterns from the book, I started writing a ‘Python Patterns’ web site that would explain the problems with the old patterns and then show some Pythonic alternatives. But that left me with a question: if we don't use Design Patterns, then what is the shape of the code we create today? In this talk we will step back and look at three kinds of thinking that are all at work together when we write software, and how our journey into learning Python prepares us for each one. false https://pretalx.com/pycon-lithuania-2025/talk/3RPFYL/ https://pretalx.com/pycon-lithuania-2025/talk/3RPFYL/feedback/ 203 Talk 2025-04-23T11:00:00+03:00 11:00 00:25 Architecture as Code (AaC) was born for prototyping a new system architecture design without any design tools. Available tools currently support on-premise and main major providers including AWS, Azure, and GCP cloud platforms. pycon-lithuania-2025-65791-architecture-as-code-aac-with-python Python Day - Apr 23 Ruslan Korniichuk en Have you already learned the benefits of the Infrastructure as Code (IaC) process? Not bad, now it’s time for the Architecture as Code (AaC) process. No worries, Python code only, no more JSON or YAML. AaC was born for prototyping a new system architecture design without any design tools. Available tools currently support on-premise and main major providers including AWS, Azure, and GCP cloud platforms. In addition, AaC allows you to track the architecture diagram changes in any version control system. false https://pretalx.com/pycon-lithuania-2025/talk/DLUSUQ/ https://pretalx.com/pycon-lithuania-2025/talk/DLUSUQ/feedback/ 203 Talk 2025-04-23T11:30:00+03:00 11:30 00:25 Explore the transformative journey of Python's Global Interpreter Lock (GIL). Delve into the GIL's origins, its role in Python's growth, and its challenges for multicore processing in development. Let's discover the implications of its experimental removal in Python 3.13.1, and how this shift might redefine concurrency, performance, and the future landscape of Python applications across various domains pycon-lithuania-2025-63261-beyond-the-gil-python-s-evolution-and-future-directions Python Day - Apr 23 /media/pycon-lithuania-2025/submissions/HU7CCA/GIL_PyTalk_INKvHPG.png Vladas Tamošaitis en false https://pretalx.com/pycon-lithuania-2025/talk/HU7CCA/ https://pretalx.com/pycon-lithuania-2025/talk/HU7CCA/feedback/ 203 Talk 2025-04-23T12:00:00+03:00 12:00 00:25 Can you imagine a python project without any return? Is it overhead or memory saving? Is it complicated or would it reduce complexity? Is it testable or a horror for unittesting? pycon-lithuania-2025-64733-the-art-of-yield Python Day - Apr 23 /media/pycon-lithuania-2025/submissions/DLW9M3/photo_2025-02-1_mYZw8FU.jpg Maxim Danilov en The general idea of this talk is: to effectively use generators in code, we need to change our programming style, and it's not too easy, but it's possible. During my talk I will convert functions and methods from the project into generators, and we will see what is effective, what is not, and where it is still better to use retur This talk is a quintessence of experience in python projects built exclusively on generators. Generators don't speed up the project directly, but they open up the possibility of lazy data processing, which in turn reduces memory consumption and can improve the performance of the project in general. Thanks to changes in recent versions of Python, the advantages of generators over common functions are becoming significant in every modern high-load data consumption project. false https://pretalx.com/pycon-lithuania-2025/talk/DLW9M3/ https://pretalx.com/pycon-lithuania-2025/talk/DLW9M3/feedback/ 203 Talk 2025-04-23T13:30:00+03:00 13:30 00:25 The batch mechanism is challenging when handling continuous data migration with DataProc. However, I'm introducing a new approach for continuous data pipelines enabled by PySpark. The participants will learn new methods to handle data consistency and reserve data completeness in a million-scale migration from SQL database into NoSQL, MongoDB. pycon-lithuania-2025-61090-migrating-billions-records-from-sql-to-nosql-using-continuous-migration-technique-with-pyspark-and-dataproc Python Day - Apr 23 Piti Champeethong en In this talk, I'll present the challenging journey in the real world from my real-world use cases to migrate millions of rows of data from SQL database into NoSQL, MongoDB. The talks composes of: - Business context and technical challenge of million rows data migration. - Data Pipeline Architecture -> SQL Server, GCP DataProc, GCP BigQuery, PySpark, and MongoDB Atlas. - Suggesting approach for handle million-row migration for SQL to NoSQL MongoDB false https://pretalx.com/pycon-lithuania-2025/talk/X7BDBK/ https://pretalx.com/pycon-lithuania-2025/talk/X7BDBK/feedback/ 203 Talk 2025-04-23T14:00:00+03:00 14:00 00:25 Business specifications are often vague or incomplete, making development challenging. Acceptance Test-Driven Development (ATDD) bridges this gap with clear, executable specifications. This talk explores how Robot Framework enhances collaboration between business and development teams. Through practical examples, we’ll show how to write effective tests and extend Robot Framework with custom Python libraries. Gain insights and tools to improve communication, development, and software delivery. pycon-lithuania-2025-65738-let-the-robots-test-acceptance-test-driven-development-atdd-with-robot-framework Python Day - Apr 23 Stefan Kraus en Developers often face the challenge of working with unclear, ambiguous, or abstract requirements from business stakeholders, leading to misaligned expectations, wasted effort, and costly rework. Without a shared understanding, delivering the right product becomes difficult, resulting in frustration and delays. Acceptance Test-Driven Development (ATDD) addresses this by providing clear, testable specifications that align both business and development teams. ATDD turns requirements into concrete, executable examples, creating a common "source of truth" that reduces misunderstandings and builds confidence in the software being developed. Robot Framework is a simple yet powerful tool for implementing ATDD, offering a keyword-driven, human-readable syntax that enables seamless collaboration between technical and non-technical stakeholders. This makes specifications accessible and executable for everyone, from business analysts to developers. In this talk, we’ll explore the core concepts of Robot Framework, showing how tests are written in its clear syntax and how existing libraries simplify common test automation tasks. We’ll also demonstrate how to extend Robot Framework with custom Python libraries to address specific project needs. You’ll leave with practical insights on using Robot Framework to build collaborative, maintainable, and effective test suites that enhance team alignment and software quality. false https://pretalx.com/pycon-lithuania-2025/talk/VELBPK/ https://pretalx.com/pycon-lithuania-2025/talk/VELBPK/feedback/ 203 Talk 2025-04-23T14:30:00+03:00 14:30 00:25 Discover how Protobuf can transform your REST API's schema evolution, while offering performance gains over JSON. This session covers Protobuf's strong versioning, ensuring seamless API updates without breaking clients. We'll tackle the challenges we faced at KAYAK, like the learning curve and integration complexity, offering strategies to address them. Gain practical insights and benchmarks as we discuss integrating Protobuf with Python frameworks, boosting your API's efficiency and adaptability. pycon-lithuania-2025-64713-transforming-rest-apis-with-protobuf-unlocking-performance-and-flexibility Python Day - Apr 23 Davi Nascimento de Paula en In this session we explore how KAYAK adopted Protobuf with REST APIs, replacing the traditional usage of JSON. Designed for API developers, software architects, and Python enthusiasts, this talk will equip you with the knowledge to elevate your API strategies with cutting-edge technology. We'll begin by exploring Protobuf's robust versioning capabilities, which allow for seamless API updates without breaking existing clients. This ensures compatibility and facilitates smooth transitions as your API evolves. But the benefits of Protobuf extend beyond versioning. One of its standout features is the significant performance gains it offers over JSON. By using a binary format, Protobuf reduces the size of data payloads, leading to faster serialization and deserialization processes. This efficiency boost can greatly enhance your API's responsiveness and scalability, making it a compelling choice for high-performance applications. Our session is packed with practical insights. We'll share real-world examples and benchmarks that highlight the tangible benefits of integrating Protobuf with Python frameworks. You'll see firsthand how Protobuf can optimize your API's performance, transforming it into a more efficient and adaptable service. Of course, adopting new technology comes with challenges. We'll address common issues such as the learning curve and integration complexities, providing actionable strategies to navigate these hurdles. Our goal is to share with you the path we adopted at KAYAK to successfully implement Protobuf in our projects. false https://pretalx.com/pycon-lithuania-2025/talk/LD8NME/ https://pretalx.com/pycon-lithuania-2025/talk/LD8NME/feedback/ 203 Talk 2025-04-23T15:00:00+03:00 15:00 00:25 Virtual environments are a fundamental part of Python development, but to most developers, they’re largely a ‘black box’. In this talk, we’re gonna dissect the code, file structure and utilities that make them up to deeply learn, and not just have superficial knowledge of, how venvs actually work. pycon-lithuania-2025-61737-inside-the-black-box-the-anatomy-of-virtual-environments Python Day - Apr 23 /media/pycon-lithuania-2025/submissions/YVWSTP/venv_anatomy_uLQrkOE.jpg Daniel Hervás en We all use virtual environments, but do we know how they work? What’s inside a virtual environment? Why do we even have to ‘activate’ them anyways, and what does that mean in the first place? In this talk, we’re gonna discover that. First, we’ll need to go over how Python installs work in modern operating systems (hopefully, the days of xkcd 1987 https://xkcd.com/1987/ are long gone), and how we can’t isolate projects without the aid of a tool such as these virtual environments. Then, we’ll go over the files and folders inside venvs, examine their purpose, and browse through the venv module source code to understand the CLI tool we use every day. We’ll install some deps and see how those fit inside this virtual environment structure. Lastly, we’ll check which problems are not solved by virtual environments, its limitations depending on the scope we’re looking for (i.e, a venv is not a container! And shouldn’t be used as such!), and how, lately, many libraries which purpose overlap with those of venvs are actually wrapping over this functionality so that you don’t have to manage it yourself (beware the law of leaky abstractions!). false https://pretalx.com/pycon-lithuania-2025/talk/YVWSTP/ https://pretalx.com/pycon-lithuania-2025/talk/YVWSTP/feedback/ 228 Workshop 2025-04-23T11:00:00+03:00 11:00 01:25 Let us build on examples from NumPy, pandas, and Matplotlib to explore techniques and tools with the Sphinx documentation generator. Learn how to implement styles, include advanced elements, and overcome challenges in creating clear, maintainable docs. 📑✨ pycon-lithuania-2025-63337-what-we-can-learn-from-exemplary-python-documentation Python Day - Apr 23 /media/pycon-lithuania-2025/submissions/YXPRYA/session_image_QqtsrVf.jpg Christian Heitzmann en If you’ve attended one of last year’s Python conferences in Europe, you might have seen my talk “Documenting Python Code”, where I introduced attendees to the basics of Python documentation. This year, I will expand on that foundation by looking at what can be learned from exemplary Python documentation. Building on renowned examples from popular Python libraries such as NumPy, pandas, and Matplotlib, this workshop will delve into techniques and tools that help streamline documentation creation while improving clarity and usability. The topics covered include: Sphinx documentation generator • use reStructuredText as a markup language • simplify docstrings with the NumPy and Google format • generate function and method documentation with sphinx-apidoc How to • include code snippets as examples and display them in environment-specific tabs • adopt themes and implement appealing styles • use admonitions to highlight important content • write sophisticated mathematical formulas • create versatile diagrams with Kroki • generate interactive HTML documentation with embedded notebooks Comparison: AsciiDoc vs. Sphinx • explore their strengths and limitations for multi-language projects • see how AsciiDoc can document APIs by including tagged source code snippets This workshop provides practical insights and examples to help developers, technical writers, and maintainers create better documentation and improve their workflows. Whether you are just starting out or refining an established project, this session will provide actionable techniques to overcome common challenges and take your documentation to the next level. You will need a Python IDE (preferable PyCharm) with a fresh project. In its terminal, please execute: pip install --upgrade pip setuptools pip install jupyterlite-pyodide-kernel pip install jupyterlite-sphinx pip install matplotlib pip install numpy pip install pandas pip install pydata-sphinx-theme pip install sphinx-book-theme pip install sphinx-copybutton pip install sphinx- false https://pretalx.com/pycon-lithuania-2025/talk/YXPRYA/ https://pretalx.com/pycon-lithuania-2025/talk/YXPRYA/feedback/ 228 Workshop 2025-04-23T14:00:00+03:00 14:00 00:55 The Creator Developer Economy offers developers the chance to turn their skills into a passive income stream. In this talk, I’ll explore how developers can leverage Apify's tools to build, deploy, and monetize web scraping solutions. From using Crawlee for Python to create efficient scrapers to publishing and earning. pycon-lithuania-2025-59088-build-deploy-monetize-the-future-of-the-developer-economy Python Day - Apr 23 Matej Hamaš en The Creator Developer Economy offers developers the chance to turn their skills into a passive income stream. In this talk, I’ll explore how developers can leverage Apify's tools to build, deploy, and monetize web scraping solutions. From using Crawlee for Python to create efficient scrapers to publishing and earning. This session will showcase the journey of building and monetizing data extraction tools in 4 steps: 1. [Development](https://docs.apify.com/platform/actors/development). 2. [Publication](https://docs.apify.com/platform/actors/publishing/publish) and set up of [monetization](https://docs.apify.com/platform/actors/publishing/monetize). 3. [Testing](https://docs.apify.com/platform/actors/development/automated-tests). 4. [Promotion](https://docs.apify.com/academy/get-most-of-actors/seo-and-promotion). This will be followed by a live demo of Apify’s CLI—a tool to quickly build and deploy Actors. We’ll walk through various types of scrapers, or "Actors" as we call them, that can be created. I’ll share tips on optimizing these tools for monetization and explain how to engage with the developer community on Discord for support. To wrap up, I’ll share real success stories from developers who earn thousands of dollars monthly by using Apify's platform, highlighting the potential of this ecosystem. false https://pretalx.com/pycon-lithuania-2025/talk/LAG8AJ/ https://pretalx.com/pycon-lithuania-2025/talk/LAG8AJ/feedback/ 228 Talk 2025-04-23T15:00:00+03:00 15:00 00:25 After creating a great web app using Python such as with flask, the next hurdle to production is how to make it available to users and operate it. And not just your app, but also ingress, the database, observability and the list goes on. We will go through your options for simplifying the operations of your web app using open source tooling. This will include using k8s directly with helm charts, PaaS using fly.io and new tooling developed by Canonical using juju. By the end of the talk you will have seen th pycon-lithuania-2025-58543-i-want-to-deploy-my-flask-app-on-kubernetes-what-are-my-options Python Day - Apr 23 David AnderssonJavier de la Puente en false https://pretalx.com/pycon-lithuania-2025/talk/3D9L9A/ https://pretalx.com/pycon-lithuania-2025/talk/3D9L9A/feedback/ 218 Workshops Talk 2025-04-23T11:00:00+03:00 11:00 00:45 Python has emerged as a versatile tool for 3D computer graphics, offering powerful capabilities in modeling, animation, and simulation. This presentation explores the application of Python in creating dynamic and visually engaging 3D graphics using Blender. The session will showcase practical examples that demonstrate Python's potential in various aspects of 3D graphics pycon-lithuania-2025-65636-python-in-3d-computer-graphics Python Day - Apr 23 Jurgis Zagorskas en Explore techniques for programmatically creating and manipulating 3D models in Blender using Python scripting. Illustrate how Python can be utilized to animate statistical data in a 3D environment. Present a Python implementation of the Game of Life cellular automaton within Blender. Showcase Python scripts for creating randomized animations in Blender. Through these demonstrations, attendees will gain insights into Python's role as a powerful tool for enhancing creativity and productivity in 3D computer graphics. The presentation aims to inspire participants to leverage Python's capabilities for innovative and expressive 3D visualizations across various domains. false https://pretalx.com/pycon-lithuania-2025/talk/HWQ3H7/ https://pretalx.com/pycon-lithuania-2025/talk/HWQ3H7/feedback/ 218 Workshops Workshop 2025-04-23T14:00:00+03:00 14:00 03:00 The new version of Django has several important features that allow us to avoid installing additional modules. Libraries like DRF, yasg, spectacular have always been recommended for REST API development. Now the rules have changed. pycon-lithuania-2025-64734-building-pure-django-rest-api Python Day - Apr 23 /media/pycon-lithuania-2025/submissions/P3BTS9/Building_pure_D_a2YugMP.jpg Maxim Danilov en In this workshop: 1. The modern art of project development in Django. (µ-Django style). 2. Built-in serializer tools. 3. Json response, an underrated tool. 4. Async ORM and class based views . The best parts of modern Django. After this workshop, all participants got information about the new features of Django. false https://pretalx.com/pycon-lithuania-2025/talk/P3BTS9/ https://pretalx.com/pycon-lithuania-2025/talk/P3BTS9/feedback/ 101 Talk 2025-04-24T09:00:00+03:00 09:00 00:25 A pycon-lithuania-2025-68684-data-day-opening Org Tomas Peluritis en false https://pretalx.com/pycon-lithuania-2025/talk/ZQATT3/ https://pretalx.com/pycon-lithuania-2025/talk/ZQATT3/feedback/ 101 Keynote 2025-04-24T09:30:00+03:00 09:30 01:00 . pycon-lithuania-2025-66157-build-your-own-simple-static-code-analyzer Keynote Stefanie Molin en Code reviewers often face significant cognitive load. Depending on the project, they must scrutinize the implementation, check that the code adheres to conventions – such as using the latest syntax and language constructs – verify that the code is properly documented, and much more. When performed manually, these code-quality checks can easily monopolize the reviewer’s time. As a result, it is a key priority to offload as many tasks as possible onto static code analysis tools – like linters and formatters – so the reviewer can focus on the implementation itself. How do these tools work, and how could you build one of your own to enforce conventions specific to your codebase? In this keynote, I will walk you through the process of creating a simple static code analyzer in Python using a data structure called an abstract syntax tree, which represents your code's structure and allows you to access its components in order to perform checks. false https://pretalx.com/pycon-lithuania-2025/talk/RAXDT9/ https://pretalx.com/pycon-lithuania-2025/talk/RAXDT9/feedback/ 101 Talk 2025-04-24T11:00:00+03:00 11:00 00:25 Many organizations have migrated their data warehouses to datalake solutions in recent years. With the convergence of the data warehouse and the data lake, a new data management paradigm has emerged that combines the best of 2 approaches: the botton-up of big data and the top-down of a classic data warehouse. pycon-lithuania-2025-61618-data-warehouses-meet-data-lakes Data Day - Apr 24 Mauro Pelucchi en In this talk, I will explain the current challenges of a datalake and how we can approach a moderm data architecture with the help of pyspark, hudi, delta.io or iceberg. We will see how organize data in a data lake to support real-time processing of applications and analyzes across all varieties of data sets, structured and unstructured, how provides the scale needed to support enterprise-wide digital transformation and creates one unique source of data for multiple audiences. false https://pretalx.com/pycon-lithuania-2025/talk/PA9K8G/ https://pretalx.com/pycon-lithuania-2025/talk/PA9K8G/feedback/ 101 Talk 2025-04-24T11:30:00+03:00 11:30 00:25 Maintaining Business Intelligent Tool (BI) governance, managing permissions, syncing documentation, and handling schema changes, can be chaotic. This talk explores how Python, Pydantic, and smart design patterns automate these tasks, ensuring seamless BI tool governance. Learn how to auto-sync table metadata, adjust queries on column renames, and enforce permissions effortlessly. With real-world examples, discover how to transform BI maintenance from a headache into a streamlined, automated process. pycon-lithuania-2025-64732-from-chaos-to-control-automating-bi-tools-with-pydantic-and-python Data Day - Apr 24 /media/pycon-lithuania-2025/submissions/BQNXNS/Automating_BI_T_Kwp0bPj.png Patricia Goldberg en Managing Business Intelligence (BI) tools at scale can quickly become chaotic. Permissions must be enforced, documentation must stay up to date, and queries must be maintained, especially when schema changes occur. Without automation, these tasks become tedious, error-prone, and time-consuming. In this talk, we’ll explore how to bring order to BI governance using Python, Pydantic, and effective design patterns. We’ll dive into three key automation strategies: 1. Automatic Permission System: Ensure users have access without manual intervention. 2. Documentation Synchronization: Keep table and column metadata up to date effortlessly. 3. Query Adjustments on Schema Changes: Prevent breaking queries by dynamically handling column renames. Using real-world examples, we’ll explain how Python’s type validation and structured data models help enforce consistency, and how design patterns streamline these processes. Whether you're struggling with BI governance or looking for inspiration to optimize your data infrastructure, this session will provide actionable insights to help you shift from chaos to control. false https://pretalx.com/pycon-lithuania-2025/talk/BQNXNS/ https://pretalx.com/pycon-lithuania-2025/talk/BQNXNS/feedback/ 101 Talk 2025-04-24T12:00:00+03:00 12:00 00:25 Real-time data analytics is essential for powering modern applications like monitoring, personalization, search, and to some extend, RAG pipelines. However, building systems that can handle real-time ingestion, indexing, and retrieval at scale is no trivial task. This talk provides actionable insights into designing and maintaining such systems at scale using best practices. pycon-lithuania-2025-62940-real-time-data-analytics-at-scale-from-ingestion-to-retrieval Data Day - Apr 24 Tung Hoang en Real-time data analytics systems are the backbone of modern applications, from monitoring and personalization to powering search experiences. However, building scalable systems that handle real-time ingestion, indexing, and retrieval efficiently can be daunting. This talk will explore the key components of such systems, including ingestion pipelines (Apache Kafka, Apache Pulsar etc), indexing layers (Elasticsearch, OpenSearch, FAISS, etc), and computation engines (Apache Flink, Apache Spark, custom Python-based solutions...). With the arrival of LLMs and Retrieval Augmented Generation (RAG), data indexation and retrieval become more critical than ever. We will address complex scenarios such as *disaster recovery, data reconciliation, and ensuring low-latency performance at scale*. Attendees will leave with actionable insights and best practices to design, implement, and maintain robust real-time analytics systems tailored to their needs. false https://pretalx.com/pycon-lithuania-2025/talk/XSGGJJ/ https://pretalx.com/pycon-lithuania-2025/talk/XSGGJJ/feedback/ 101 Talk 2025-04-24T14:00:00+03:00 14:00 00:25 Our mission is simple but profound: to improve and extend lives by learning from the experience of every person with cancer. This talk explains how we transform sensitive data from heterogeneous environments into research-grade datasets. And how we shift insights generation left to iterate faster. pycon-lithuania-2025-65802-accelerating-privacy-enhancing-data-processing Data Day - Apr 24 Florian Stefan en In this talk, we’ll begin by introducing the concept of real-world evidence datasets and their transformative impact on cancer research. We’ll explore the significant challenges of building high-quality real-world evidence datasets, including the fragmented healthcare data landscape, the complexity of source data, and stringent regulatory constraints. To address these challenges, we’ll introduce our privacy-enhancing data architecture and foundational technology stack, which includes AWS, Snowflake, Python, DLT, DBT, Pandas, and DuckDB. We’ll explain how we leverage these tools to establish critical feedback loops and accelerate actionable insights. Key topics include: - Integrating notebooks with production pipelines to bridge the gap between clinical domain expertise and production workflows. - Employing a write-audit-publish pattern to ensure continuous data quality through unit tests, data tests, regression tests, and cohort-spanning statistical tests. - Using DLT to move normalization steps closer to the data source, unlocking iterative data modeling and enabling more agile development cycles. One of our most impactful innovations was shifting data investigations to the left. Previously, raw data inspections were delayed until pre-processed data was available in our data warehouse, causing prolonged feedback loops and inefficiencies. Disorganized file formats often required data scientists to manually inspect data, limiting their ability to draw meaningful, cohort-wide conclusions. To overcome these challenges, we established local databases directly on compute instances where raw data is stored, leveraging the flexibility and transparency of this approach. Transformations developed within this setup can seamlessly transfer to the cloud data warehouse while remaining portable and adaptable to various environments. This enabled our data scientists to explore and understand the data earlier in the process, eliminating bottlenecks. With immediate access to organized datasets, our tea false https://pretalx.com/pycon-lithuania-2025/talk/8JTKYP/ https://pretalx.com/pycon-lithuania-2025/talk/8JTKYP/feedback/ 101 Talk 2025-04-24T14:30:00+03:00 14:30 00:25 In data science, speed matters as much as accuracy, especially when users expect quick results. This talk explores simple yet effective techniques to boost performance and responsiveness on data-centric web apps based on practical experience working with Panel apps. While some strategies are case-specific, most apply broadly to data-driven projects. pycon-lithuania-2025-65737-working-for-a-faster-world-accelerating-data-science-with-less-resources Data Day - Apr 24 Maximilian Lattka en Performance is critical in data science — accuracy alone isn’t enough if applications are slow. Users expect both correct and fast results, and delays can lead to frustration, decreased productivity, and reduced trust in the tools. Whether in web apps, dashboards, or data pipelines, efficient processing is essential for user satisfaction and business success. This talk explores simple yet effective techniques to optimize performance, focusing on faster data processing and analysis. Based on past experience working with Panel to create data-centric web apps, we demonstrate strategies such as non-blocking processing, streaming or lazy-loading to reduce bottlenecks and enhance speed. While some methods are case-specific, most are broadly applicable across data science projects. These scalable solutions improve data handling and computation, enabling faster insights and real-time decision-making. By implementing these strategies, teams can significantly reduce latency and enhance user experience without compromising accuracy. Whether working on web apps, internal tools, or large-scale pipelines, these approaches help tackle performance challenges effectively. false https://pretalx.com/pycon-lithuania-2025/talk/7JGQBN/ https://pretalx.com/pycon-lithuania-2025/talk/7JGQBN/feedback/ 101 Talk 2025-04-24T15:00:00+03:00 15:00 00:25 Hear more about the evolving landscape of SQL transformation tools and data lineage challenges. Explore how sqlglot enables powerful SQL parsing and transformation capabilities, and see practical demonstrations of sqlmesh as a modern alternative to dbt. Learn about open-source approaches to data lineage tracking and discover how these tools are shaping the future of data engineering workflows. pycon-lithuania-2025-60062-beyond-dbt-modern-sql-transformation-and-lineage-with-sqlglot-and-sqlmesh Data Day - Apr 24 Tomas Peluritis en This talk explores the intersection of SQL transformation frameworks and data lineage tracking, focusing on open-source solutions that are changing how we handle data transformations at scale. We'll begin by examining common pain points in data lineage tracking, particularly when dealing with complex SQL transformations across different dialects and platforms. The first part will deep dive into sqlglot's architecture and demonstrate how it serves as a crucial building block for modern data tools by enabling dialect-agnostic SQL parsing, analysis, and transformation. We'll explore real-world use cases where sqlglot's capabilities unlock new possibilities for data lineage tracking and SQL optimization. Next, we'll contrast sqlmesh with dbt, highlighting key architectural differences and their implications for data engineering workflows. Through live demonstrations, we'll showcase sqlmesh's unique features including time-travel capabilities, automated dependency management, and built-in data lineage tracking. We'll also address how these tools approach column-level lineage tracking, comparing open-source alternatives to proprietary solutions like dbt Cloud. The session will conclude with practical guidelines for implementing these tools in your data stack and a discussion of future trends in SQL transformation and lineage tracking. Whether you're a data engineer looking to optimize your workflow or an architect evaluating data transformation frameworks, you'll leave with actionable insights about modern SQL tooling. false https://pretalx.com/pycon-lithuania-2025/talk/FH39NT/ https://pretalx.com/pycon-lithuania-2025/talk/FH39NT/feedback/ 101 Talk 2025-04-24T15:30:00+03:00 15:30 00:25 A successful data scientist needs to have solid coding skills and stay up to date with the latest artificial intelligence and machine learning algorithms. However, there are many other skills and experiences that help you succeed in data science. In this talk Megan shares five of her most helpful career lessons she's learned in over eight years as a data scientist. These lessons will include tips on advocating for your own career development, how to collaborate with other teams and more. pycon-lithuania-2025-65815-top-5-lessons-from-a-senior-data-scientist Data Day - Apr 24 Megan Robertson en After attending this talk the audience will walk away with tips and strategies for performing better in their career. While not directly related to Python, it is applicable for those who write code for their roles and are in technical positions. I will share five of the most important lessons I have learned in my 8 year career as a data scientist that have helped me earn promotions and achieve success at work. You can find a brief summary of the some of the lessons below. It is important to know how to advocate for yourself and make sure that your contributions are noticed - this talk will discuss how to tactfully share your accomplishments. In addition the talk will go over how to collaborate with non-technical teams effectively who do not share the same coding background. I will also discuss some of the approaches I take to writing code and developing models when I start a new project. false https://pretalx.com/pycon-lithuania-2025/talk/C9D9C9/ https://pretalx.com/pycon-lithuania-2025/talk/C9D9C9/feedback/ 101 Keynote 2025-04-24T16:30:00+03:00 16:30 01:00 Data management systems have gone through significant changes in the last 10 years, driven by user demands, novel techniques and improvements in hardware. These have far-reaching implications on how systems are deployed and used in practice. In this talk, I will focus on three key aspects of modern data management systems: scalability, mutability, and interface. I will share my personal experiences, and will bring several examples from the database and data science worlds. pycon-lithuania-2025-66952-the-evolution-of-data-management-techniques Keynote Gabor Szarnyas en false https://pretalx.com/pycon-lithuania-2025/talk/LXVE7A/ https://pretalx.com/pycon-lithuania-2025/talk/LXVE7A/feedback/ 203 Talk 2025-04-24T11:00:00+03:00 11:00 00:25 Machine learning models are never truly “done.” As data evolves, so should the models that rely on it. But how can we ensure continuous improvement without costly retraining or manual intervention? In this talk, we introduce an automated pipeline designed to incrementally enhance model performance by systematically testing and integrating new features. pycon-lithuania-2025-67173-beyond-deployment-continuously-adding-features-to-drive-marginal-gains-in-models Data Day - Apr 24 Mark Fukson en Machine learning models are never truly “done.” As data evolves, so should the models that rely on it. But how can we ensure continuous improvement without costly retraining or manual intervention? In this talk, we introduce an automated pipeline designed to incrementally enhance model performance by systematically testing and integrating new features. Using Deep Feature Synthesis (DFS), we generate a dynamic pool of candidate features and evaluate their impact on predictive power. Only features that demonstrably improve the model are added, ensuring continuous refinement without unnecessary complexity. This process transforms model performance monitoring from a passive task into an active, value-driven strategy. Attendees will learn: - How to automate feature selection and model enhancement at scale. - The role of DFS in discovering new predictive signals. - Best practices for integrating incremental feature addition into production workflows. true https://pretalx.com/pycon-lithuania-2025/talk/GFRRFN/ https://pretalx.com/pycon-lithuania-2025/talk/GFRRFN/feedback/ 203 Talk 2025-04-24T11:30:00+03:00 11:30 00:25 The "data build tool" (DBT) was designed to unlock software engineering best practices for SQL-based data pipelines: pipelines as version controlled directed acyclic graphs (DAGs) consisting of testable and reusable nodes. With the increasing number of cloud data warehouses and data lakehouses that allow the native execution of Python code, DBT also added support for Python models. In this talk, I will explain how Flatiron Health uses DBT and share our experiences with unit and data testing. pycon-lithuania-2025-68732-testable-data-pipelines Data Day - Apr 24 Florian Stefan en false https://pretalx.com/pycon-lithuania-2025/talk/DNQ3L8/ https://pretalx.com/pycon-lithuania-2025/talk/DNQ3L8/feedback/ 203 Talk 2025-04-24T12:00:00+03:00 12:00 00:45 In this talk, we introduce cluster-experiments, a Python library designed to facilitate end-to-end A/B testing workflows, including power analysis, experiment analysis, and variance reduction techniques. pycon-lithuania-2025-65740-cluster-experiments-a-python-library-for-end-to-end-a-b-testing-workflows Data Day - Apr 24 David Masip en We will go through the main techniques for mde analysis (simulation based and using Central limit theorem), how to do variance reduction in mde analysis and how to analyse experiments with the same library. false https://pretalx.com/pycon-lithuania-2025/talk/PKXYNE/ https://pretalx.com/pycon-lithuania-2025/talk/PKXYNE/feedback/ 203 Talk 2025-04-24T14:00:00+03:00 14:00 00:25 Good retrieval performance is key to an effective RAG system, as it ensures relevant information is selected, directly impacting augmentation and generation quality. My presentation focuses on RAG indexing and retrieval, exploring methods to convert text into searchable formats, comparing techniques, and analyzing their advantages, disadvantages, and performance on an annotated dataset to enhance document retrieval based on user queries. pycon-lithuania-2025-66097-smarter-retrieval-better-generation-improving-rag-systems Data Day - Apr 24 David Batista en Retrieval Augmented Generation (RAG) is a model architecture for tasks requiring information retrieval from large corpora combined with generative models to fulfill a user information need. It's typically used for question-answering, fact-checking, summarization, and information discovery. The RAG process consists of indexing, which converts textual data into searchable formats; retrieval, which selects relevant documents for a query using different methods; and augmentation, which feeds retrieved information and the user's query into a Large Language Model (LLM) via a prompt for output generation. Typically, one has little control over the augmentation step besides what's provided to the LLM via the prompt and a few parameters, like the maximum length of the generated text or the temperature of the sampling process. On the other hand, the indexing and retrieval steps are more flexible and can be customized to the specific needs of the task or the data. My talk will focus on RAG systems' indexing and retrieval techniques. Attendees will learn about various methods, starting with classical approaches rooted in the Information Retrieval community. While these methods have been around for decades, they remain widely used today due to their simplicity and efficiency. The session will then explore more modern techniques that leverage LLMs to enhance the indexing process or optimize user queries. These approaches aim to improve the retrieval of relevant documents and improve the performance of RAG systems. Participants will gain insights into each technique's unique features, advantages, and limitations, along with guidance on selecting the most appropriate approach for specific tasks and datasets. The talk will conclude with a performance analysis, comparing the implementation of all these techniques in Python using Haystack and evaluating them over an annotated dataset. Speed, accuracy, and efficiency will be assessed, offering an understanding of trade-offs and practical takeaways. false https://pretalx.com/pycon-lithuania-2025/talk/S83YCX/ https://pretalx.com/pycon-lithuania-2025/talk/S83YCX/feedback/ 203 Talk 2025-04-24T14:30:00+03:00 14:30 00:25 A brag document is a powerful tool to highlight your work by making it visible, measurable, and demonstrating its real impact on you and your organisation - but such a document can be time-consuming to maintain. My talk explores automation of the writing process with language models fed with data from tools like Jira, Notion, and code commits. Learn how to save time, avoid registering missed achievements, and make your work stand out. Ideal for engineers at all levels looking to grow their impact. pycon-lithuania-2025-65800-automate-brag-document-writing-with-llms Data Day - Apr 24 Ludvig Wärnberg Gerdin en Getting your work noticed is often the key to expanding your impact and moving forward in your career. It’s not just about doing great work—it’s about making sure it’s being seen and recognized. I recognised this first hand in my previous roles, both as a manager but also as an individual contributor. That’s where brag documents come in. First introduced by Julie Evans and championed by leaders like Will Larson (author of Staff Engineer, An Elegant Puzzle), they’re a simple but powerful way to keep track of your achievements, allowing you to showcase your work and your manager to make a clear promotion case for you. The problem? They can be tedious to maintain, leading many of us to put them off and forget key wins. I will explore and show how ingesting data from task management tools (Jira, Notion, Linear) and code commits, summarising the information with language models and storing the summaries, you keep an up-to-date brag doc. I’ll walk you through the framework, how it works, how I use it to keep my work visible and aligned with goals, and how you can use it. This session is valuable for python engineers at all levels who want to make their impact more visible, grow their impact, and advance their career. The tone is informative and no prior knowledge required —just curiosity as well a desire to grow and to optimise their workflow. **Outline:** - Why visibility matters: Growing your impact through recognition. - The brag document basics: How and why they work. - How to automate that process with language models: A look at how it pulls data from your daily tools and makes use of it. - My workflow: How I use it to save time and boost visibility. - Key takeaways: Key takeaways from the talk false https://pretalx.com/pycon-lithuania-2025/talk/3QPMF8/ https://pretalx.com/pycon-lithuania-2025/talk/3QPMF8/feedback/ 203 Talk 2025-04-24T15:00:00+03:00 15:00 00:25 This presentation examines approaches for detecting and eliminating near-duplicate images across datasets ranging from small collections to repositories containing millions of images. We will compare the performance of several embedding models, including CLIP, ResNet, and other variants, assessing their ability to capture semantic and perceptual similarity and performance tradeoffs. We will benchmark various vector database solutions on query speed and memory consumption. pycon-lithuania-2025-66163-image-deduplication-using-embeddings Data Day - Apr 24 Jonas Jarutis en false https://pretalx.com/pycon-lithuania-2025/talk/W37VCC/ https://pretalx.com/pycon-lithuania-2025/talk/W37VCC/feedback/ 203 Talk 2025-04-24T15:30:00+03:00 15:30 00:25 Variable selection is often left up to an algorithm. However, controlling for some variables can improve measurement accuracy, and thus overall performance. On the other hand, certain "bad" controls can block pathways of relationships between variables that we want to preserve or create spurious correlations. Using real and simulated data, I explain when to reconsider your controls, and why that may significantly improve model accuracy. pycon-lithuania-2025-62755-variable-selection-what-your-model-can-t-tell-you Data Day - Apr 24 James Donahue en Econometricians spend a lot of time thinking about causality, whereas data science generally focuses more on prediction and classification. But is there something to be learned from economists' fixation on causal relationships? Variable selection is often left up to an algorithm. However, controlling for some variables can improve measurement accuracy, and thus overall performance. On the other hand, certain "bad" controls can block pathways of relationships between variables that we want to preserve or create spurious correlations. Using real and simulated data, I explain when to reconsider controls, and why that may significantly improve model accuracy. false https://pretalx.com/pycon-lithuania-2025/talk/D3QNLM/ https://pretalx.com/pycon-lithuania-2025/talk/D3QNLM/feedback/ 228 Talk 2025-04-24T11:00:00+03:00 11:00 00:25 In this talk, I’ll share how I began trading stocks and why I turned to Python to track my performance—along with the abundance of surprises that came with it. We’ll walk through the building blocks of two Python-powered apps: one that extracts stock transactions from screenshots, and another that generates summaries of my trading to uncover valuable insights pycon-lithuania-2025-64715-how-i-tracked-my-stocks-with-python Data Day - Apr 24 Ąžuolas Krušna en What started as a curiosity about stock trading turned into a Python-powered journey of discovery. In this talk, I’ll share how I built a system to track my trades, analyze performance, and gain insights that even the trading platforms themselves didn’t offer. We’ll explore two small but powerful apps: one that extracts stock transactions directly from screenshots using OCR and regex, and another that imports CSV data to generate summaries of trading performance. Along the way, you’ll see how I designed a regex lab to iteratively refine data extraction, how I structured and stored data with DuckDB, and how I visualized everything using Streamlit—all with simplicity and reusability in mind. This isn’t just about trading. It’s about using Python to build small, personal automations that grow into tools you rely on. Whether you’re interested in OCR, regex, data analysis, or just looking for ideas to automate your own workflows, you’ll leave with something practical—and maybe inspired to track your own story in code false https://pretalx.com/pycon-lithuania-2025/talk/ND9SPX/ https://pretalx.com/pycon-lithuania-2025/talk/ND9SPX/feedback/ 228 Talk 2025-04-24T11:30:00+03:00 11:30 00:25 A case study of rewriting a simple data pipeline involving Python, a pinch of Go, Git workflows, Airflow, Postgres and Cloud. Investigating some common assumptions and principles of designing data pipelines. The benefits and issues with the tools and how these may be handled. I hope this case study of a pipeline rewrite will give you insights that are applicable to Python use for your own data pipelines, and into cloud pricing. pycon-lithuania-2025-65921-cutting-the-price-of-scraping-cloud-costs Data Day - Apr 24 /media/pycon-lithuania-2025/submissions/PZD7QE/EDBskus_MD2ATsh.png Ed Crewe en Get answers to the following options and more, as to what is the cheapest and most maintainable solution for this kind of data pipeline. 1. Do the data scraping ourselves OR use an established 3rd party aggregated data source 2. Use temporary embedded DBs OR a cloud master DB server 3. Use a standard ELT pipeline pattern OR do ETL within each pipeline step 4. Implement purely via the pipeline DAGs and SQL OR create a separate Python package The following areas will be covered. ## How the Airflow data pipeline framework can be used. The approach to the architecture and reasons for the rewrite. Python code for scraping data, along with Soda testing to verify the steps data. The client data consumption architecture to our Go cloud service ## How does cloud pricing work? It is complicated! For example, hard disk storage cost should be simple? But it varies based on hardware type, size, throughput, iops, different price depending on zone (us-east2 vs ap-south1 etc.), charges for traffic between regions vs within regions Then there are the backup costs based on its retention size/time, schedule. ## What data sources does our case study need? Cloud price lists are not small. The full combined price lists for AWS, Google and Azure are 5 million prices. Review possible data sources and related tools, such as Kubecost, Infracost.io and the cloud providers direct sources. ## What is the cost of the different pipeline architectures Often data scientists put together a PoC pipeline without considering cost, it becomes production and starts eating money! Charges for cloud SaaS are more concealed than raw Cloud PaaS. But it must be assessed too, and they have some interesting pricing. For example if your data pipeline is open source and public, Github workflows are free! If it is private they charge twice as much for compute as using direct Azure. Similarly cloud pipeline providers are big in this space, but how does Astronomer SaaS pricing stack up vs. running yourself on cloud PaaS? false https://pretalx.com/pycon-lithuania-2025/talk/PZD7QE/ https://pretalx.com/pycon-lithuania-2025/talk/PZD7QE/feedback/ 228 Talk 2025-04-24T12:00:00+03:00 12:00 00:25 We will dive into the fascinating world of football analytics, showcasing how to collect and process match data (e.g., Hudl Statsbomb, Sportmonks, and Understat), including player tracking, event logs, and tactical formations. Attendees will walk away with practical knowledge and Jupyter Notebooks, demonstrating Python's power in decoding modern football strategies. pycon-lithuania-2025-65797-python-on-the-pitch-how-germany-will-win-world-cup-2026 Data Day - Apr 24 Ruslan Korniichuk en In this talk, we will explore how Python can be leveraged to analyze and visualize football data for the [Germany national football team](https://en.wikipedia.org/wiki/Germany_national_football_team), managed by [Julian Nagelsmann](https://en.wikipedia.org/wiki/Julian_Nagelsmann), on their journey to winning the 2026 FIFA World Cup. We will dive into the fascinating world of **football analytics**, showcasing how to collect and process match data (e.g., Hudl Statsbomb, Sportmonks, and Understat), including player tracking, event logs, and tactical formations. We'll discover match data to demonstrate how the team's performance reflects Nagelsmann's tactical principles, such as [gegenpressing](https://en.wikipedia.org/wiki/Gegenpressing), offensive play, and compactness. Join us to unlock the power of Python in football analytics! false https://pretalx.com/pycon-lithuania-2025/talk/3PMUYD/ https://pretalx.com/pycon-lithuania-2025/talk/3PMUYD/feedback/ 228 Talk 2025-04-24T14:00:00+03:00 14:00 00:25 Forecasting is a common activity that has clear business value in various domains but it is not a very common skill that Data Scientists have or feel confident about. In this crash course I will cover the fundamentals of Time Series forecasting from the basic methods to more advanced techniques. I will do this showcasing practical code examples using libraries from Nixtla. pycon-lithuania-2025-65822-a-crash-course-in-time-series-forecasting-from-naive-to-foundational Data Day - Apr 24 Pietro Peterlongo en This is a technical talk that aims to provide the essential elements for a Data Scientist to go from zero knowledge of time series forecasting to being able to code a forecasting system that can evolve from a basic setup that can be put in production. After a brief introduction to motivation why we forecast and the different domains and use cases, the rest of the presentation will consists of code examples that will exemplify the various steps of a data science lifecycle. We will start from looking at the data, implementing a baseline model and set up an evaluation framework. From there we will explore different type models from statistical to ML and different techniques (probabilistic forecasting and hierarchical forecasting). Finally we will comment on more advanced time series methods like neural methods and foundational models for time series. The code examples will make use of Nixtla libraries, because we think they currently represent the state of the art for time series forecasting with Python both in term of range of functionality and consistency of API. Most of the concepts and techniques can translate to other frameworks. false https://pretalx.com/pycon-lithuania-2025/talk/NKFV38/ https://pretalx.com/pycon-lithuania-2025/talk/NKFV38/feedback/ 228 Talk 2025-04-24T14:30:00+03:00 14:30 00:25 Temporal is an open source, distributed, and scalable workflow orchestration platform designed to execute mission-critical business logic with resilience. Manage failures, network outages, flaky endpoints, long-running processes and more, ensuring your workflows never fail. pycon-lithuania-2025-65795-temporal-bulletproof-workflows Data Day - Apr 24 Ruslan Korniichuk en Are you ready for the next generation of orchestration platforms? Temporal is the successor to Uber Cadence. It is an open source, distributed, and scalable workflow orchestration platform to execute mission-critical business logic with resilience. Your code will run reliably even if it encounters problems, such as network outages or server crashes. With great Developer Experience (DX) and Python SDK, the platform simplifies coding. Join us to unlock the power of Python and Temporal in mastering workflow orchestration! false https://pretalx.com/pycon-lithuania-2025/talk/VGRBRM/ https://pretalx.com/pycon-lithuania-2025/talk/VGRBRM/feedback/ 228 Talk 2025-04-24T15:00:00+03:00 15:00 00:25 An intro to the Dagster open-source orchestration tool. Data Tool Stack. What is Dagster, and who is it for? What are its main use cases? Testing the data and the code. Deployment ideas to production. pycon-lithuania-2025-68752-organize-your-data-stack-using-dagster Data Day - Apr 24 /media/pycon-lithuania-2025/submissions/RKPMDS/Dagster_intro_uhgCRMt.png Patricia Goldberg en This talk is going to be an introduction to Dagster as an open-source orchestration tool. I'll show a high-level view of a data tool stack architecture and how Dagster integrates it all. From what assets, resources, and schedules are to how to test and deploy, let me show you use cases for this Pythonic tool. false https://pretalx.com/pycon-lithuania-2025/talk/RKPMDS/ https://pretalx.com/pycon-lithuania-2025/talk/RKPMDS/feedback/ 228 Talk 2025-04-24T15:30:00+03:00 15:30 00:25 Are you using Airflow or Pandas? Great! You've contributed to better data management at your organization. The breakthrough of AI has reignited focus on high-quality data and effective data governance (not that scary as it sounds!) and management practices. AI needs fit-for-purpose data to reach its potential, and we already have powerful toolkit — like Airflow, Pandas, Matplotlib/Seaborn, or Great Expectations — to optimize workflows and ensure data quality. pycon-lithuania-2025-65889-the-power-of-python-for-data-management-or-how-you-ve-been-doing-data-management-all-along-without-even-realizing-it Data Day - Apr 24 Vidmantė Čižienė en Are you using Airflow or Pandas? Great! You've been contributing to better data management at your organization. Now, let’s take one step back. Data governance and management professionals have eagerly awaited this moment. The breakthrough of AI has reignited the focus on high-quality data and effective data governance (it's not as scary as it sounds!) and management practices and tools. Although these disciplines aren't new, they've finally emerged from the shadow of big data, data engineering, cloud engineering and other things we’ve been focused on lately. Ultimately, AI needs fit-for-purpose data to reach its full potential. How do we obtain reliable and accessible data? Fortunately, we already have powerful tools in our tech stack to enhance our data management practices. Let's explore how Python-based solutions and libraries — specifically Airflow and Pandas, along with Matplotlib/Seaborn for visualization or Great Expectations for data quality checks, and other — can optimize our workflows and ensure data quality. By utilizing these tools, you are instrumental in driving your organization toward its AI goals and better data management. Even if you’ve never framed your work that way before. P.S. And when it comes to definitions, data governance isn’t as scary as it sounds. The most useful definition I've encountered is that data governance is really about “preventing people from doing stupid stuff with data” (Charlotte Ledoux, LinkedIn post feed). And good data management in this light, would be actually doing smart stuff with data. false https://pretalx.com/pycon-lithuania-2025/talk/HDVLAY/ https://pretalx.com/pycon-lithuania-2025/talk/HDVLAY/feedback/ 218 Workshops Workshop 2025-04-24T11:00:00+03:00 11:00 00:55 This talk explores the synergy between Apache Beam and Apache Airflow, demonstrating how to create a robust, end-to-end data engineering workflow. We'll dive into the challenges of orchestrating complex data processing tasks and show how combining Airflow's scheduling capabilities with Beam's data processing framework can create more efficient and manageable data pipelines. The session will cover integration with Google Cloud Platform services, including Cloud Functions, BigQuery, and Gemini AI models. pycon-lithuania-2025-65798-orchestrating-an-end-to-end-data-engineering-workflow-leveraging-python-in-apache-beam-and-airflow Data Day - Apr 24 Sadeeq Akintola en Problem Addressed: In today's data-driven world, organizations face the daunting challenge of orchestrating complex, end-to-end data engineering workflows that seamlessly integrate batch and streaming processing, scheduling, cloud services, and AI models. This talk tackles the often-overlooked synergy between Apache Beam and Apache Airflow, two powerful tools in the data engineering ecosystem that are rarely used in tandem. We'll explore how combining these technologies with Google Cloud Platform services and cutting-edge AI models can revolutionize data pipeline architecture. Relevance to the Audience: As data volumes explode and processing requirements become increasingly complex, data engineers and scientists are under pressure to build scalable, maintainable pipelines that can handle diverse data sources and downstream applications. This topic is crucial for professionals looking to: - Modernize their data infrastructure - Streamline machine learning pipelines - Overcome limitations of using Beam and Airflow separately - Integrate AI models into data workflows seamlessly - Leverage cloud services for enhanced scalability and performance Solutions and Key Takeaways: Attendees will gain practical insights and hands-on knowledge to: 1. Harness the Power of Integration: Learn to seamlessly combine Apache Beam's robust data processing capabilities with Apache Airflow's sophisticated scheduling and orchestration features. 2. Master Cloud-Native Data Engineering: Discover how to leverage Google Cloud Platform services like Cloud Functions and BigQuery to build serverless, scalable data pipelines. 3. Incorporate AI into Data Workflows: Explore techniques for integrating Gemini AI models into your data processing pipelines, opening new possibilities for intelligent data transformation and analysis. 4. Design Resilient Architectures: Gain expertise in creating modular, scalable, and fault-tolerant data architectures that can handle the demands of modern data-driven applications. false https://pretalx.com/pycon-lithuania-2025/talk/MZ8DBC/ https://pretalx.com/pycon-lithuania-2025/talk/MZ8DBC/feedback/ 218 Workshops Workshop 2025-04-24T12:00:00+03:00 12:00 00:55 Do you ever find it complicated to learn the complexities of a traditional web framework to push your data science work online? Worry no more! Streamlit might help speed things up as it is designed for the required purpose - creating beautiful data-related web apps that can be deployed in minutes. In the hands-on tutorial, we’ll go through various features of Streamlit and build a small lyric fetcher app based on the available curated dataset of around 24K Billboard top-100 songs. pycon-lithuania-2025-65841-build-deploy-apps-like-a-pro-data-scientist-using-streamlit Data Day - Apr 24 Siddharth Gupta en 0:01-0:05 minutes: In the first section, I will discuss with you the basics of Streamlit and some examples of applications made through it. I will also show you the expected final version of what we’ll create during the tutorial. 00:05 - 0:15 minutes: In the second section, I will run a small “Hello World” code on the local server, to give you the initial feel of streamlit. 00:15 - 0:50 minutes: In the third section, I will build our application step-by-step by creating a layout and adding the required elements. These elements would include two drop-down buttons for selecting the song & artist for which we want lyrics, A lyric showcasing column, and a word cloud visualization of the respective lyrics. In the last 5 minutes, I will touch on how we can deploy the app online using Heroku and Streamlit, which you can further attempt after the talk on your own. false https://pretalx.com/pycon-lithuania-2025/talk/JYPT3E/ https://pretalx.com/pycon-lithuania-2025/talk/JYPT3E/feedback/ 218 Workshops Workshop 2025-04-24T14:00:00+03:00 14:00 00:55 In today’s fast-paced machine learning environment, the ability to efficiently manage and reuse features across multiple models is crucial. This workshop explores how leveraging a feature store can streamline ML pipelines by ensuring consistency and accelerating deployment cycles. Participants will gain hands-on experience with setting up, managing, and integrating feature stores into their existing workflows—transforming raw data into valuable, production-ready features. pycon-lithuania-2025-64414-using-feature-stores-to-deliver-awesome-models Data Day - Apr 24 Laurynas StašysMantas Cepulkovskis en Join us for an immersive, hands-on session designed for data scientists, ML engineers, and AI enthusiasts eager to optimize their machine learning pipelines through advanced feature store functionality. In this workshop, we will focus on the robust capabilities of an open-source feature store platform Feast that streamlines the entire feature lifecycle without requiring you to reinvent the wheel. What to Expect: • Unified Data Management: Learn how to consolidate offline and online feature data into a single, cohesive system. Discover strategies to ensure that the same high-quality features used during training are available during inference, eliminating training-serving skew and improving model consistency. • Scalable Feature Engineering: Explore methods to automate data transformations and store reusable components. This session will show you how to reduce redundancy and accelerate model iteration by centralizing your feature definitions. • Hands-On Integration: You will learn how to deploy Feast feature store into Kubernetes. By the end of the workshop, you’ll have practical experience and actionable insights to implement a feature store that elevates your machine learning workflows in Kubernetes environment. false https://pretalx.com/pycon-lithuania-2025/talk/LQTZB3/ https://pretalx.com/pycon-lithuania-2025/talk/LQTZB3/feedback/ 218 Workshops Workshop 2025-04-24T15:00:00+03:00 15:00 00:55 We will explore the landscape of technical analysis libraries available for the Python language, including popular choices like TA-Lib (aka talib), Pandas TA, and Technical Analysis (aka bukosabino/ta) library. pycon-lithuania-2025-65793-investing-technical-analysis-libraries-in-python Data Day - Apr 24 Ruslan Korniichuk en Do you have financial savings to invest?! In this 55-minute workshop, we'll explore the landscape of technical analysis libraries available for Python, including popular choices like TA-Lib (aka talib), Pandas TA, and Technical Analysis (aka bukosabino/ta) library. We'll explore their capabilities, comparing their pros and cons. Join us to unlock the power of Python in mastering technical analysis! false https://pretalx.com/pycon-lithuania-2025/talk/NVHMDY/ https://pretalx.com/pycon-lithuania-2025/talk/NVHMDY/feedback/ 101 Talk 2025-04-25T09:00:00+03:00 09:00 00:25 A pycon-lithuania-2025-68686-ai-day-opening Org en false https://pretalx.com/pycon-lithuania-2025/talk/ZNHAXV/ https://pretalx.com/pycon-lithuania-2025/talk/ZNHAXV/feedback/ 101 Keynote 2025-04-25T09:30:00+03:00 09:30 01:00 Multimodal AI is booming this year with models capable of seeing, reading, hearing. Models advancing in this field unlocks many production use cases in robotics, document AI, computer/web automations and more! In this talk we will go through everything multimodal and open-source: a bit of background, libraries, very basic APIs to get you started with open-source models, popular open-source models, use cases (multimodal agents, multimodal RAG, automated browser use and more!) pycon-lithuania-2025-66813-open-source-multimodal-ai Keynote Merve Noyan en false https://pretalx.com/pycon-lithuania-2025/talk/SSR3JJ/ https://pretalx.com/pycon-lithuania-2025/talk/SSR3JJ/feedback/ 101 Talk 2025-04-25T11:00:00+03:00 11:00 00:25 Some of the latest big evolutionary steps in generative AI has been models that support function calling and “agentic” capabilities. This is provides generative models with “tools” that allow them to go beyond generating outputs for simple queries, and start planning the best way to solve complex queries. In this talk, we’ll be diving into using vector databases as the backbone for these types of complex AI architectures, both serving as knowledge bases, and memory. pycon-lithuania-2025-62004-knowledge-bases-memory-for-agentic-ai AI Day - Apr 25 Tuana Çelik en In this technical talk, we will start by covering the history of how agentic AI came about. We will go over how we can design prompts in a way that instruct LLMs to use tools and plan out how to solve complex queries. Next, we will learn about function calling and how this feature of LLMs can be used as the basis of agents. The talk will include a small amount of coding, ending in a working agent in the form of a technical assistant. false https://pretalx.com/pycon-lithuania-2025/talk/VTUXX8/ https://pretalx.com/pycon-lithuania-2025/talk/VTUXX8/feedback/ 101 Talk 2025-04-25T11:30:00+03:00 11:30 00:25 Discover how EGTL (Extract, Generate, Transfer, Load) extends traditional ETL by adding a “generate” step powered by GenAI. In this talk, I’ll demonstrate how Python pipelines on top of data warehouse can automatically extract data, generate new insights, and deliver optimized transformations. We’ll explore practical workflows, real-world use cases, and best practices—equipping you to apply EGTL in your own data projects. pycon-lithuania-2025-61507-egtl-data-processing-model-prototype-using-python AI Day - Apr 25 Aleksejs Vesjolijs en Experience a novel approach to data ingestion pipelines with EGTL (Extract, Generate, Transfer, Load), a practical evolution of the standard ETL process. By incorporating a “generate” step powered by GenAI into the workflow, EGTL unlocks new potential in data transformation, advanced data analytics, and automation. In this talk, I’ll walk you through the core components of EGTL, demonstrating how Python-based data pipelines can leverage generative AI to produce enriched datasets on the fly before transferring and loading them into downstream systems. Using real-world use cases, I’ll illustrate how EGTL benefits both data engineers and scientists by reducing manual overhead, accelerating iterative development, and unveiling unexpected insights. You’ll learn implementation best practices—covering everything from tool selection and architecture design to error handling and governance. This talk will highlight key strategies for integrating GenAI directly into your pipelines. false https://pretalx.com/pycon-lithuania-2025/talk/YDETJ8/ https://pretalx.com/pycon-lithuania-2025/talk/YDETJ8/feedback/ 101 Talk 2025-04-25T12:00:00+03:00 12:00 00:25 This talk charts the evolution of Artificial Intelligence through the dual lenses of data and models, tracing AI’s journey from early symbolic systems to today’s advanced data-driven techniques. Attendees will learn how the interplay of ever-growing datasets and increasingly sophisticated model architectures has powered major breakthroughs, transforming AI from theoretical curiosity to a global catalyst for innovation. pycon-lithuania-2025-65886-ai-360-from-theory-to-transformation AI Day - Apr 25 Stefan Dayneko en We will begin by examining AI’s early days, when handcrafted rules and symbolic reasoning took center stage despite limited data and computational resources. Next, the spotlight shifts to the rise of machine learning and neural networks, as larger datasets and improved computing power enabled more flexible, adaptive models. Along the way, we will highlight pivotal milestones—such as the resurgence of deep learning—that propelled AI forward at an accelerated pace. By focusing on how evolving data availability and model complexity shaped each phase of AI, this session provides a comprehensive historical perspective and offers insights into how AI’s trajectory continues to unfold in today’s data-rich world. false https://pretalx.com/pycon-lithuania-2025/talk/YYLYQS/ https://pretalx.com/pycon-lithuania-2025/talk/YYLYQS/feedback/ 101 Talk 2025-04-25T14:00:00+03:00 14:00 00:25 AI-driven code generation can transform software development in regulated sectors like banking and insurance - but only if implemented securely and responsibly. In this talk, we’ll explore how to harness tools like GitHub Copilot and ChatGPT to boost productivity while ensuring compliance. Attendees will learn key considerations, best practices, and practical insights to keep code generation both efficient and fully auditable. pycon-lithuania-2025-65628-code-generation-in-regulated-industries-opportunities-and-challenges AI Day - Apr 25 /media/pycon-lithuania-2025/submissions/XUDKBG/github_2BqwmXA.png Antanas Daujotis en In highly regulated industries, code quality, security, and compliance are paramount. Yet recent advances in AI-driven coding assistants promise faster development, fewer errors, and improved agility. How do we tap into these benefits without introducing unacceptable risk? This session will: Demystify AI Code Generation: - Brief overview of tools like GitHub Copilot, Code Llama, ChatGPT and other - How these models learn, generate code, and support developers Tackle Regulatory and Compliance Challenges: - Why finance, insurance, and other sectors have stringent requirements - Ensuring auditability, explainability, and data privacy - Real-world pitfalls: potential for code vulnerabilities or data leaks Adopt Best Practices for Secure Development: - Human-in-the-loop code review and automated scanning - Model fine-tuning with domain-specific datasets - Integrating AI generation into CI/CD pipelines with robust security checks Explore Future Possibilities: - How “regulation-aware” AI might evolve - Potential for industry-specific large language models - Balancing innovation and compliance false https://pretalx.com/pycon-lithuania-2025/talk/XUDKBG/ https://pretalx.com/pycon-lithuania-2025/talk/XUDKBG/feedback/ 101 Talk 2025-04-25T14:30:00+03:00 14:30 00:25 What is the way from prove of concept to big production solutions for GenAI application? How to make it scalable and make one release by 7 sprints? pycon-lithuania-2025-65789-genai-for-clients-no-pain-no-gain AI Day - Apr 25 Daria Lashkevich en We talk about different GenAI Accenture cases in production and preproduction studies, how to implement new technologies and what advantages and disadvantages we can discover during such projects. How often client really know what they want? GenAI "magic" tricks or 99% accuracy? Let's talk about LLM, stable diffusion and similarity search. false https://pretalx.com/pycon-lithuania-2025/talk/Z8SHV8/ https://pretalx.com/pycon-lithuania-2025/talk/Z8SHV8/feedback/ 101 Talk 2025-04-25T15:00:00+03:00 15:00 00:25 We built a cutting-edge speech-to-text model that outperformed solutions from industry leaders like Microsoft, Google, and OpenAI. This is the story of how we identified a market gap, defined what makes "the best model," and turned our vision into a successful business. pycon-lithuania-2025-61604-how-we-outperformed-microsoft-google-and-openai-in-speech-to-text AI Day - Apr 25 /media/pycon-lithuania-2025/submissions/PGJM98/bitmap_R71Acg0.png Alius Petraška en In this talk, I’ll share my personal journey of how we started our own business and built an innovative audio (speech-to-text) model that outperformed the offerings of industry giants like Microsoft, Google, and OpenAI. I’ll discuss how we identified a crucial gap in the market, carried out in-depth analysis to define what makes "the best model," and overcame the challenges of competing with established players. This story is about determination, creativity, and the power of a focused vision, showing that even small teams can achieve big results when they find the right opportunities. false https://pretalx.com/pycon-lithuania-2025/talk/PGJM98/ https://pretalx.com/pycon-lithuania-2025/talk/PGJM98/feedback/ 101 Talk 2025-04-25T15:30:00+03:00 15:30 00:25 Financial transactions generate vast amounts of sequential data, yet traditional risk assessment models often rely on predefined features that may not capture the full complexity of user behavior. This talk explores how transaction embeddings—inspired by techniques from NLP and Computer Vision—can transform financial modeling. pycon-lithuania-2025-67269-from-raw-transactions-to-insights-the-power-of-embeddings-in-fintech AI Day - Apr 25 Hanna Danilovich en In financial services, transactional data holds valuable insights, yet traditional models often rely on feature engineering, which can be time-consuming and restrictive. This talk explores how transaction embeddings can capture richer representations of financial behavior, leading to more accurate and scalable models. We’ll discuss how these embeddings are generated, the techniques used to learn meaningful representations, and their integration into machine learning models to enhance predictive performance. Attendees will gain practical insights into leveraging embeddings in fintech applications, with a focus on improving risk estimation and decision-making. false https://pretalx.com/pycon-lithuania-2025/talk/HKJC3N/ https://pretalx.com/pycon-lithuania-2025/talk/HKJC3N/feedback/ 101 Lightning talk session 2025-04-25T16:15:00+03:00 16:15 00:15 ⚡ pycon-lithuania-2025-69029-lightning-talks Lightning talks en false https://pretalx.com/pycon-lithuania-2025/talk/VN9EZZ/ https://pretalx.com/pycon-lithuania-2025/talk/VN9EZZ/feedback/ 101 Keynote 2025-04-25T16:30:00+03:00 16:30 01:00 2025 is positioned to be the year of Agentic Systems - AI agents capable of autonomous decision-making that are transforming the software landscape. In just a few months, we have already seen the technology transition through multiple hype cycles. In this talk, we’ll cut through the noise to explore the real challenges and emerging opportunities in the space. We will examine where we are, where we're headed, and what it all means for developers shaping the future of AI. pycon-lithuania-2025-67012-challenges-and-opportunities-of-agentic-systems-present-and-future Keynote Aurimas Griciunas en false https://pretalx.com/pycon-lithuania-2025/talk/DACKBS/ https://pretalx.com/pycon-lithuania-2025/talk/DACKBS/feedback/ 203 Talk 2025-04-25T11:00:00+03:00 11:00 00:25 Pydantic, Pancakes and Poter quick though deep dive into the world of response modeling pycon-lithuania-2025-65709-pydantic-crafting-confidence-in-ai-apps AI Day - Apr 25 Gediminas Sadaunykas en Pydantic has become a cornerstone of AI development over the past few years, serving as a critical component in major SDKs including OpenAI, Anthropic, LangChain, and CrewAI. Through tools like Instructor (and later Structured Outputs), Pydantic has enabled the deployment of apps enhanced via LLMs in production environments. Now, Pydantic AI is extending this capability to AI Agents. As AI communicates through data, Pydantic's data validation framework has become the internal language for AI systems. Lets' meet to explore fun and fascinating world of response modeling! false https://pretalx.com/pycon-lithuania-2025/talk/8N8JCE/ https://pretalx.com/pycon-lithuania-2025/talk/8N8JCE/feedback/ 203 Talk 2025-04-25T11:30:00+03:00 11:30 00:25 In an era dominated by data, businesses struggle with processing diverse, unstructured information across systems. This research presents an AI-powered pipeline addressing product matching challenges in retail and e-commerce. Our solution combines traditional matching algorithms with deep learning through a five-step process. This approach minimizes manual intervention while improving accuracy and efficiency. pycon-lithuania-2025-62893-the-best-of-both-worlds-a-hybrid-approach-to-lightning-fast-product-matching AI Day - Apr 25 Zafarzhon Irismetov en In today's data-driven world, businesses face the challenge of managing vast amounts of information, often from diverse and unstructured sources. This complexity necessitates efficient and accurate data processing techniques to extract meaningful insights and drive informed decision-making. One such critical task is product matching, where the goal is to accurately identify and link records representing the same product across different systems or datasets. This is particularly crucial for businesses operating in industries with complex product catalogs, such as retail, manufacturing, and e-commerce. Our project solves this problem by developing a robust, AI-powered pipeline to automate the matching process while minimizing human intervention. The task involved matching incoming product records with a standardized catalog, referred to as the "G_List," which includes essential attributes. The primary challenge lay in the inherent variability and inconsistency of product data. Incoming records often included: spelling and grammatical errors, data inconsistencies, multilingual variations. Examples of the challenge include identifying that "Yellow Glass whiskey 1L," "Color Whiskey 1L glas," and "GLAS amarillo whiskey" all correspond to "GLASS YELLOW WHISKEY 1L" in the G_List. Our solution comprises five key steps. First, we created the G_List by standardizing product data and attributes. Next, we consolidated source data into a unified data lake, where records were cleaned and transformed. The third step utilized the tfidf_matcher Python library for initial matching. Step four, feature extraction. Finally, a deep learning model assessed the matches as either "PASS" or "NOT PASS." Records flagged as "PASS" were automatically labeled and finalized, while "NOT PASS" records were escalated for manual review. This hybrid approach of combining matching algorithms with AI significantly reduced manual workload while improving accuracy. false https://pretalx.com/pycon-lithuania-2025/talk/QVW7WA/ https://pretalx.com/pycon-lithuania-2025/talk/QVW7WA/feedback/ 203 Talk 2025-04-25T12:00:00+03:00 12:00 00:25 AI agents are transforming the way we create applications. However, developing multi-agent applications can often feel complex and time-consuming. LangFlow simplifies this process by offering an intuitive, easy-to-use interface for building AI-driven solutions. pycon-lithuania-2025-65733-building-multi-agent-ai-applications-made-easy-with-langflow AI Day - Apr 25 Christophe Bornet en AI agents are transforming the way we create applications. However, developing multi-agent applications can often feel complex and time-consuming. LangFlow, the leading Python-based and open-source visual editor for low-code AI applications with over 45k stars on GitHub, simplifies this process by offering an intuitive, easy-to-use interface for building AI-driven solutions. In this talk, we will start by introducing the concept of AI agents—what they are, how they work, and why they’re important for modern applications. Then, we will dive into LangFlow, showcasing how its visual editor allows you to quickly create powerful AI applications without needing to write extensive code. Finally, we’ll bring everything together with a live demo, where we’ll build a multi-agent application from scratch, step by step. By the end of the session, you’ll have a clear understanding of how LangFlow can help you build efficient, multi-agent AI systems with ease, and you’ll be ready to start using it to create your own intelligent applications. false https://pretalx.com/pycon-lithuania-2025/talk/TS88N3/ https://pretalx.com/pycon-lithuania-2025/talk/TS88N3/feedback/ 203 Talk 2025-04-25T14:00:00+03:00 14:00 00:25 How can machine learning enhance biomedical image analysis? This talk explores the potential of Python and PyTorch in automating artifact and damage segmentation. From data preprocessing to clustering-based label classification and deep learning-driven segmentation, key techniques will be discussed, including the use of Convolutional Neural Network architectures. The session will also cover performance evaluation and insights into advancing biomedical imaging with AI-driven solutions. pycon-lithuania-2025-64387-unlocking-the-power-of-python-and-pytorch-for-biomedical-image-segmentation AI Day - Apr 25 Taisija Kozarina en The field of pathology is undergoing rapid digitalization, integrating new technologies that optimize workflows, reduce manual effort, and enhance diagnostic accuracy. However, preprocessing remains a labor-intensive step, requiring specialists to manually segment out artifacts before further analysis can take place. This talk explores how Python and PyTorch can be leveraged to automate this crucial preprocessing stage using machine learning techniques. The session will begin with an overview of the challenges in biomedical image segmentation, emphasizing the importance of artifact detection in digital pathology. It will then explore machine learning-based approaches, particularly the role of convolutional neural networks (CNNs) in automating segmentation. Special focus will be given to U-Net and YOLO architectures, discussing their efficiency in handling biomedical image data. Additionally, the talk will cover label classification using clustering algorithms such as K-Means and DBSCAN, which help refine annotation quality and improve segmentation outcomes. The discussion will extend to model training, evaluation, and the comparison of different architectures based on performance metrics. Practical insights from experimental results will be shared, including challenges encountered during development and potential solutions. The session will conclude by addressing the broader impact of AI-powered segmentation in digital pathology, discussing future advancements, ethical considerations, and recommendations for further research. Attendees will gain a deeper understanding of how Python and PyTorch can streamline biomedical image preprocessing, reduce manual workload, and enhance the accuracy of pathological analysis through automation. Keywords: Semantic segmentation, Computer Vision, Convolutional Neural Network, Clustering, Semi-supervised Learning, Digital Pathology, Machine Learning with Python false https://pretalx.com/pycon-lithuania-2025/talk/BMXAPA/ https://pretalx.com/pycon-lithuania-2025/talk/BMXAPA/feedback/ 203 Talk 2025-04-25T14:30:00+03:00 14:30 00:25 Unlock sensitive data potential with anonymization! Learn how Python, diffusion models, and Named Entity Recognition (NER) empower institutions to anonymize PII in financial documents, replacing it with synthetic stand-ins. Discover open-source, self-hosted tools to ensure privacy while unleashing data's full power. pycon-lithuania-2025-61190-anonymization-of-sensitive-information-in-financial-documents-using-python-diffusion-models-and-named-entity-recognition AI Day - Apr 25 /media/pycon-lithuania-2025/submissions/WG9MMX/anonymization_W3uCGY2.jpg Piotr Gryko en Anonymization of sensitive information in financial documents using, python, diffusion models and named entity recognition Data is the fossil fuel of the machine learning world, essential for developing high quality models but in limited supply. Yet institutions handling sensitive documents — such as financial, medical, or legal records often cannot fully leverage their own data due to stringent privacy, compliance, and security requirements, making training high quality models difficult. A promising solution is to replace the personally identifiable information (PII) with realistic synthetic stand-ins, whilst leaving the rest of the document in tact. In this talk, we will discuss the use of open source tools and models that can be self hosted to anonymize documents. We will go over the various approaches for Named Entity Recognition (NER) to identify sensitive entities and the use of diffusion models to inpaint anonymized content. false https://pretalx.com/pycon-lithuania-2025/talk/WG9MMX/ https://pretalx.com/pycon-lithuania-2025/talk/WG9MMX/feedback/ 203 Talk 2025-04-25T15:00:00+03:00 15:00 00:25 Fairness and safety are fundamental criteria for building trustworthy and high-quality AI systems, whether they are credit scoring models, hiring assistants, or healthcare chatbots. But what does it truly mean for an AI system to be fair and safe? In this talk, I will explore the potential risks and challenges associated with these principles and introduce various approaches and techniques for evaluating AI systems. The discussion will center on applications powered by Large Language Models (LLMs). pycon-lithuania-2025-65782-how-to-evaluate-fairness-and-safety-in-llm-applications AI Day - Apr 25 Sebastian Krauss en Artificial intelligence holds immense promise to revolutionize industries and enhance lives, but its true potential hinges on trust. Fairness and safety are not just ethical ideals—they are critical pillars for the responsible development, deployment, and evaluation of AI systems. This is particularly important for applications powered by Large Language Models (LLMs), which are now shaping diverse fields like coaching, hiring, and financial services. However, ensuring these models meet rigorous fairness and safety standards presents complex challenges. In this 25-minute talk, we will dive into five key areas: 1. Introduction to LLM-Based Applications: We’ll begin with an accessible introduction to Large Language Models—avoiding technical deep dives - and showcase real-world examples of how LLMs are applied across industries. 2. Understanding Risks and Challenges: LLMs, while powerful, come with vulnerabilities such as biases and the potential for harmful behaviors. Discriminatory or toxic outputs can infringe on human rights, damage reputations, and result in financial loss. Through real-world examples, I will highlight these risks and discuss why addressing them is crucial. 3. Defining Fairness and Safety in AI: What does it mean for an AI system to be fair and safe? Here, I will outline the key criteria for trustworthy AI applications, offering a framework to evaluate their ethical and societal impact. 4. Evaluation Techniques and Tools: From public benchmarks to innovative prompt designs and metrics, we’ll explore practical methods for assessing LLM trustworthiness. This section will empower participants to leverage the right tools to ensure their AI systems meet ethical and operational standards. 5. Building the Future: Strategies for Fair and Safe AI: The final segment focuses on actionable strategies to design and deploy fair, safe, and compliant AI systems. I’ll also discuss the role of emerging regulations, such as the EU AI Act, in guiding responsible innovation. false https://pretalx.com/pycon-lithuania-2025/talk/UHVHWG/ https://pretalx.com/pycon-lithuania-2025/talk/UHVHWG/feedback/ 203 Talk 2025-04-25T15:30:00+03:00 15:30 00:25 Modern work demands constant context-switching—emails, notes, meetings, and tasks pile up, leaving us overwhelmed. This talk introduces a <b>slow productivity AI</b> approach, inspired by <b>Cal Newport</b>, that leverages <b>offline, open-source automation</b> using <b>Hugging Face, n8n, and Obsidian</b>. By structuring knowledge into meaningful tasks <b>without disrupting deep work</b>, we can create a <b>sustainable, low-distraction workflow</b>—working smarter, not just faster. pycon-lithuania-2025-66134-slow-productivity-ai-automating-knowledge-task-management-with-offline-hugging-face-n8n-obsidian AI Day - Apr 25 /media/pycon-lithuania-2025/submissions/DHLHQ3/ai-slow-product_DDVD9wp.png piotr stepinski en In his book Slow Productivity, Cal Newport argues that modern work culture prioritizes busyness over effectiveness, leading to stress, shallow work, and burnout. But what if AI could enhance deep work rather than create more digital noise? This talk explores AI for slow productivity, leveraging open-source automation to reduce cognitive overload, structure knowledge, and enhance focus—while remaining fully private and offline. <b>The Problem</b> Knowledge workers juggle vast amounts of unstructured information: - Scattered Notes (Obsidian, meeting transcripts, handwritten thoughts) - Endless Emails (buried action items in Gmail/Outlook) - Disjointed Task Lists (manual tracking in GitHub Issues, Todoist, Notion) - Unstructured Schedules (missed follow-ups, unclear priorities) Instead of chasing hyper-productivity, we embrace Cal Newport’s slow productivity principles: ✅ Work at a natural pace—Automate routine tasks without adding friction. ✅ Prioritize meaningful work—AI helps extract what truly matters from information chaos. ✅ Reduce distractions—A fully offline, self-hosted workflow supports deep work. <b>The Solution: AI-Powered, Private Slow Productivity </b> This talk introduces a fully offline, open-source automation pipeline for structuring knowledge and task management: - Hugging Face Transformers (Local NLP for Meaning-Making) – Extract insights from raw notes, emails, and transcripts. - n8n (Self-Hosted Automation) – Connect data sources, enabling automation <b>without third-party cloud services.</b> - Obsidian + GitHub Issues – Convert scattered knowledge into structured tasks <b>without disrupting deep focus.</b> - Google Calendar / Teams (Self-Hosted Sync) – Automate scheduling while respecting slow productivity. false https://pretalx.com/pycon-lithuania-2025/talk/DHLHQ3/ https://pretalx.com/pycon-lithuania-2025/talk/DHLHQ3/feedback/ 228 Talk 2025-04-25T11:30:00+03:00 11:30 00:25 AI is evolving from passive tools to autonomous agents, driving the rise of agentic workflows that can plan, execute, and optimize tasks with minimal human input. This session will explore how large language models and multi-agent collaboration power these systems, enhancing efficiency and innovation. Attendees will gain insights into real-world applications, challenges, and the future of autonomous AI systems, uncovering how agentic workflows are transforming industries. pycon-lithuania-2025-64601-the-emergence-of-agentic-workflows-in-ai AI Day - Apr 25 /media/pycon-lithuania-2025/submissions/HH9ACL/_d5548901-75eb_H6AOljT.jpeg Robert Dzisevič en As AI systems transition from passive tools to autonomous agents, agentic workflows are emerging as a new paradigm. These AI-driven systems autonomously plan, execute, and optimize tasks with minimal human input, transforming areas like automation and decision-making. This session will focus on how developers can leverage large language models and multi-agent collaboration to build efficient, innovative agentic workflows. Code examples using a popular agentic framework will be demonstrated, providing hands-on insights into real-world applications. Attendees will also explore challenges and future directions for autonomous AI systems, gaining valuable knowledge on how agentic workflows are accelerating AI-driven change. false https://pretalx.com/pycon-lithuania-2025/talk/HH9ACL/ https://pretalx.com/pycon-lithuania-2025/talk/HH9ACL/feedback/ 228 Talk 2025-04-25T12:00:00+03:00 12:00 00:25 What can go wrong with tokenizer encodings? Everything! I will share my experience of understanding, misunderstanding, and ultimately learning to work with tokenization in LLMs. I will discuss what surprisal is, its relevance to my research, and its connection to tokenization. The talk will include various examples illustrating how misunderstandings of tokenization can arise, as well as strategies for debugging and preventing these issues. pycon-lithuania-2025-65834-surprisal-and-the-headache-of-tokenizer-encodings-in-llms AI Day - Apr 25 /media/pycon-lithuania-2025/submissions/LXA8K7/ezgif-36b43cc87_BvqlA8G.jpg Siddharth Gupta en false https://pretalx.com/pycon-lithuania-2025/talk/LXA8K7/ https://pretalx.com/pycon-lithuania-2025/talk/LXA8K7/feedback/ 228 Talk 2025-04-25T14:00:00+03:00 14:00 00:45 Scoris.lt utilized Large Language Models (LLMs) to address the challenge of improving SEO performance by generating financial descriptions for Lithuanian companies. The case study highlights the innovative application of LLMs and custom translation models to create high-quality, multilingual content at scale. pycon-lithuania-2025-59748-leveraging-large-language-models-for-automated-generation-and-validation-of-financial-descriptions-for-lithuanian-companies AI Day - Apr 25 Antanas Baltrušaitis en The session would outline the implementation of a scalable AI solution by Scoris.lt. Facing challenges with Google search rankings due to a lack of text-based content, Scoris employed LLMs to generate financial descriptions in Lithuanian. The solution combined open-source models for cost efficiency and high throughput, supported by fine-tuned translation models for linguistic accuracy. The results included generating readable text for over 115,000 companies, boosting impressions and average rankings on Google search. This case study not only demonstrates the business impact of leveraging AI-driven tools but also offers a roadmap for integrating affordable, high-performance LLM and translation technologies. false https://pretalx.com/pycon-lithuania-2025/talk/9F3LPG/ https://pretalx.com/pycon-lithuania-2025/talk/9F3LPG/feedback/ 228 Talk 2025-04-25T15:00:00+03:00 15:00 00:25 The Model Context Protocol (MCP) is an emerging standard that enables structured data provisioning for LLMs and AI agents. However, the current data discovery mechanism in MCP is static. This limits the AI’s ability to dynamically assess the utility, relevance, and efficiency of data tool calls in real time. Here I present an enhancement to MCP "tool discovery" that introduces dynamic data descriptions, allowing LLM to be better informed. pycon-lithuania-2025-65731-enhancing-model-context-protocol-mcp-with-dynamic-tool-discovery-for-smarter-ai AI Day - Apr 25 Viraj Sharma en This talk will introduce the concept of MCP with a demonstration of how MCP clients and servers. How they are built and how they interact. Then I will focus on a specific flow - the tool discovery. It is possible to enhance the specification of the MCP allowing tools to update their metadata periodically based on real-time system and environmental factors. This enhancement will enable AI models to intelligently choose tools based on: -Data freshness (last update, data volume, change frequency) -System load & latency (server utilization, estimated response time) -API rate limits & costs (quota usage, request cost) -Geographical & time-based relevance (regional availability, peak usage) -Data accuracy & trustworthiness (confidence scores, bias detection) These are highlighted to encourage the audience to think of enhance the MCP in more ways. false Slides https://pretalx.com/pycon-lithuania-2025/talk/WRGLCJ/ https://pretalx.com/pycon-lithuania-2025/talk/WRGLCJ/feedback/ 228 Talk 2025-04-25T15:30:00+03:00 15:30 00:25 The more users your platform attracts, the more unwanted attention you'll get from people looking to game your system. While every product is unique, the journey of tackling these bad actors tends to follow similar patterns across companies. In this talk, I'll walk you through the three stages of platform protection that I've witnessed firsthand, and how to level up your safety game using the data you have. pycon-lithuania-2025-65963-when-safety-is-non-negotiable-3-stages-of-building-safety-using-data-ai AI Day - Apr 25 Evaldas Kazlauskis en Building effective platform safety isn't a sprint—it's an evolution. Drawing from real-world experience, this session breaks down the three critical phases of safety maturity: from reactive monitoring ("crawl"), to proactive protection ("walk"), to sophisticated, AI-powered prevention ("run"). We'll examine what separates each stage and the exact capabilities needed to advance your safety operations. You'll discover practical approaches to transform your existing data into powerful AI-driven safety tools that shield users from harm while preserving their experience. false https://pretalx.com/pycon-lithuania-2025/talk/87ETB8/ https://pretalx.com/pycon-lithuania-2025/talk/87ETB8/feedback/