From PyAI 2026

PyAI 2026 talks

Explore conversations about Python, agents, and AI in production. Find a talk, pick a topic, or visit the channel for every recording.

29 talks

Samuel Colvin: Controlling the wild: Monty, from tool calling to computer use

Samuel Colvin · 19:40

Controlling the wild: Monty, from tool calling to computer use

Jeremiah Lowin: Build Reasonable Software

Jeremiah Lowin · 21:25

Build Reasonable Software

Armin Ronacher: Leaning In To Find Out

Armin Ronacher · 20:57

Leaning In To Find Out

Hamel Husain: The Revenge of the Data Scientist

Hamel Husain · 18:04

The Revenge of the Data Scientist

Sebastián Ramírez: What's New in FastAPI

Sebastián Ramírez · 22:28

What's New in FastAPI

Guido van Rossum, Samuel Colvin, Sebastián Ramírez & Jeremiah Lowin: Open Source in the age of AI

Guido van Rossum, Samuel Colvin, Sebastián Ramírez & Jeremiah Lowin · 37:59

Open Source in the age of AI

Pamela Fox: Improving MCP tool schemas to increase agent reliability

Pamela Fox · 21:40

Improving MCP tool schemas to increase agent reliability

Yury Selivanov: Spec as Infrastructure

Yury Selivanov · 21:57

Spec as Infrastructure

Jeff Huber: Building a powerful agentic search agent (with Pydantic AI)

Jeff Huber · 22:10

Building a powerful agentic search agent (with Pydantic AI)

Benjamin Hindman: Building AI Chat Apps (MCP ext-apps)

Benjamin Hindman · 24:48

Building AI Chat Apps (MCP ext-apps)

Zanie Blue: Do developer tools matter to agents?

Zanie Blue · 21:49

Do developer tools matter to agents?

Bradley Axen: 3 MCP Patterns That Actually Ship: Building Goose at Block

Bradley Axen · 25:08

3 MCP Patterns That Actually Ship: Building Goose at Block

Dylan Madisetti: Notebooks That Don’t Break in Production: marimo vs. Jupyter

Dylan Madisetti · 16:02

Notebooks That Don’t Break in Production: marimo vs. Jupyter

Nithin Krishnamurthi: Replacing Redis Qs with Prefect: ML Orchestration at Snorkel AI

Nithin Krishnamurthi · 13:05

Replacing Redis Qs with Prefect: ML Orchestration at Snorkel AI

Pedro Lopez: Building 600+ Connectors with AI: Airbyte’s Automated Connector Factory

Pedro Lopez · 17:03

Building 600+ Connectors with AI: Airbyte’s Automated Connector Factory

Antoni Kozelski: Logfire AI Assistant: An Agent for Your Agents

Antoni Kozelski · 7:13

Logfire AI Assistant: An Agent for Your Agents

Chang She: 10 Things I Hate About Feature Engineering for AI

Chang She · 22:27

10 Things I Hate About Feature Engineering for AI

Ciro Greco: Git for Your Data Lake: Why Agents Need Isolation and Rollback

Ciro Greco · 22:17

Git for Your Data Lake: Why Agents Need Isolation and Rollback

Indragie Karunaratne: Claudescope: A Deep Dive into Claude Code Telemetry

Indragie Karunaratne · 10:59

Claudescope: A Deep Dive into Claude Code Telemetry

Ethan Rosenthal: ML Without the YAML

Ethan Rosenthal · 16:03

ML Without the YAML

Panel discussion: AI and ML in Production

Panel discussion · 31:05

AI and ML in Production

Simba Khadder: Context Engineering 2.0: MCP, Agentic RAG & Memory

Simba Khadder · 18:31

Context Engineering 2.0: MCP, Agentic RAG & Memory

Smit Shah: Engineering Low-Latency Voice Agents

Smit Shah · 21:05

Engineering Low-Latency Voice Agents

Alex Chau: How Ramp Enables Search across Billions of Docs

Alex Chau · 14:31

How Ramp Enables Search across Billions of Docs

Matthaus Krzykowski: Building Production-Ready Multimodal ETL Pipelines

Matthaus Krzykowski · 20:30

Building Production-Ready Multimodal ETL Pipelines

Till Döhmen: Reflections on Context Engineering for MCP Servers

Till Döhmen · 21:27

Reflections on Context Engineering for MCP Servers

Hamza Tahir: We Solved Building Agents. Now What?

Hamza Tahir · 21:45

We Solved Building Agents. Now What?

Abhishyant Khare: Agent Native Engineering With Cofounder/CTO

Abhishyant Khare · 21:42

Agent Native Engineering With Cofounder/CTO

Niamh Gavin: AI Research Fails - Anecdotes from the Frontline

Niamh Gavin · 15:56

AI Research Fails - Anecdotes from the Frontline