RoboCon 2026

Building a local RAG for Robot Framework, based on the example of API testing
2026-02-10 , Building a local RAG for Robot Framework, based on the example of API testing

In this hands-on workshop, you'll fuse AI (RAG) with Robot Framework to help with API tests. We will build a system from scratch that reads OpenAPI docs to automatically generate RF code snippets and create intelligent test data for positive, negative, and edge cases. And API test is just an example. You can use the knowledge to build RAGs for any topic.
This intermediate-level session is perfect for QA engineers and developers eager to innovate their testing process. Let's automate the automation!


Imagine an intelligent assistant that reads an OpenAPI specification and helps write your Robot Framework tests. In this intensive, hands-on workshop, you won't just imagine it—you'll build it from the ground up.

This workshop takes you on a practical journey to fuse the power of Retrieval-Augmented Generation (RAG) AI models with the reliability of Robot Framework. You will leave not just with theory, but with a working prototype and the skills to revolutionize your API testing workflow.

Your Step-by-Step Learning Journey:

  1. Build the AI's Brain: The Knowledge Base
    You'll start by tackling the core problem: making technical documentation understandable to an AI. You will learn how to take a standard OpenAPI/Swagger file, intelligently break it down ("chunking"), and convert it into numerical representations ("embeddings"). You'll then store these in a vector database, creating a powerful, searchable knowledge base that forms the foundation of our system.

  2. Master Prompt Engineering for Test Generation
    With a knowledge base in place, you will learn the art of "Prompt Engineering"—crafting precise instructions for the AI. You'll move beyond simple questions to designing sophisticated prompts that command the model to query your knowledge base and generate clean, syntactically correct, and ready-to-use Robot Framework code, complete with keywords from the RequestsLibrary.

  3. Execute, Debug, and Refine
    A generated test is useless until it runs. You will take the AI-generated .robot files and execute them against a real API. This is where your testing expertise comes in. You'll learn to analyze the results, debug any issues using Robot Framework's detailed logs, and add your own critical assertions to ensure the tests are not just running, but are truly validating the API's functionality.

  4. Generate Intelligent and Diverse Test Data
    Finally, you will push the boundaries by using the AI to automate another tedious task: test data creation. You will command the model to generate a wide array of JSON payloads based on the API schema—covering positive scenarios, negative cases (e.g., missing fields, incorrect data types), and critical edge cases. You will then learn how to integrate this data into data-driven tests using Robot Framework's Test Template feature.

By the end of this workshop, you will have built a complete, end-to-end system that automates the most time-consuming parts of API test creation, freeing you to focus on high-level test strategy and exploratory testing.

Hardware requirements:
RAM: 16 GB
Free disc space: 15 GB (SSD preferred)

The participants will have to install the following before the workshop:
VSCode + RobotCode
Python (3.9+)
Robot Framework (7.2+)
Ollama
Llama 3 8B


Lessons Learned:

building RAG from ground up. Understand embeddings, writing good prompts

Categorize / Tags:

AI, RAG, API tests

Is this suitable for ..?: Intermediate RF User, Advanced RF User Describe your intended audience:

intermediate or advanced RF user that wants to increase his/hers productivity with AI toops

In-Person or Online talk/workshop?:

in-person workshop

Senior software tester with over 15 yeas of experience. Working with RF since 2018. Exploring how AI will change the way we test software.
Apart of that RF ambassador and member of the Wrobocon organization team, co-hots of Wrobocon.

My name is Agnieszka Żminkowska.

I have been working in testing for over 10 years, with almost 8 of those years at Capgemini. I have worked on projects in the public administration and telecommunications sectors, mainly for clients from Germany.

In my projects, I am responsible for test automation. I work as a Test Architect and Quality Manager.