BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//pydata-london-2026//talk//TAWYHU
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
DTSTART:20250605T000000
TZNAME:BST
TZOFFSETFROM:+0100
TZOFFSETTO:+0100
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251026T020000
RDATE:20261025T020000
TZNAME:GMT
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20260329T020000
RDATE:20270328T020000
TZNAME:BST
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:Keynote: Samuel Colvin: Pydantic Monty & Logfire: Wild LLMs\, from
  tool calling to computer use - Samuel Colvin
DTSTART;TZID=Europe/London:20260605T132000
DTEND;TZID=Europe/London:20260605T140500
DTSTAMP:20260812T145255Z
UID:pretalx-pydata-london-2026-TAWYHU@pretalx.com
DESCRIPTION:LLMs are increasingly being used to take actions\, call APIs\,
  and write code. But giving AI agents the ability to run code opens up a s
 urprisingly tricky question: how much control do you actually hand over?\n
 \nThere's a full continuum here\, from structured tool calling at one end 
 to full computer use at the other\, but most developers don't realise how 
 many interesting options live in between. That gap matters\, because the e
 xtremes both have serious trade-offs: pure tool calling is safe but sequen
 tial and limiting\, while full sandboxes or computer use are powerful but 
 complex\, slow\, and often a hard sell to enterprise security teams.\n\nTh
 is talk introduces Monty\, a minimal Python interpreter written in Rust\, 
 purpose-built for running AI-generated code safely. Unlike traditional san
 dboxing approaches that start with full access and try to lock things down
 \, Monty starts from zero and requires you to explicitly grant each capabi
 lity — meaning the LLM can only interact with the outside world through 
 functions you wrote\, control\, and can audit. It's a new paradigm: not AI
  using your tools\, but AI writing its own programs to coordinate your too
 ls.\n\nIn this talk\, you will learn how to think about the control-capabi
 lity trade-off when building AI agents\, where Monty sits on that spectrum
  and why\, and how to use it with Pydantic AI to replace sequential tool c
 alls with expressive Python — complete with a live demo traced through L
 ogfire.\n\nBasic familiarity with Python and LLM tool use is helpful but n
 ot required. No prior knowledge of Rust or sandboxing concepts needed.
LOCATION:Grand Hall 1
URL:https://pretalx.com/pydata-london-2026/talk/TAWYHU/
END:VEVENT
END:VCALENDAR