app_store_button

google_play_button

DK - Details page - Device banner - 894x1036

Nyd den ubegrænsede adgang til tusindvis af spændende e- og lydbøger - helt gratis

Lyt når som helst, hvor som helst 

Michael Kennedy

Talk Python to Me is a weekly podcast hosted by developer and entrepreneur Michael Kennedy. We dive
 deep into the popular packages and software developers, data scientists, and incredible hobbyists doing
 amazing things with Python. If you're new to Python, you'll quickly learn the ins and outs of the community
 by hearing from the leaders. And if you've been Pythoning for years, you'll learn about your favorite
 packages and the hot new ones coming out of open source.

Talk Python To Me

In this episode, I’m talking with Vincent Warmerdam about treating LLMs as just another API in your Python app, with clear boundaries, small focused endpoints, and good monitoring. We’ll dig into patterns for wrapping these calls, caching and inspecting responses, and deciding where an LLM API actually earns its keep in your architecture.

Episode sponsors

Seer: AI Debugging, Code TALKPYTHON

NordStellar

Talk Python Courses

Links from the show
Vincent on X: @fishnets88

Vincent on Mastodon: @koaning

LLM Building Blocks for Python Co-urse: training.talkpython.fm

Top Talk Python Episodes of 2024: talkpython.fm

LLM Usage - Datasette: llm.datasette.io

DiskCache - Disk Backed Cache (Documentation): grantjenks.com

smartfunc - Turn docstrings into LLM-functions: github.com

Ollama: ollama.com

LM Studio - Local AI: lmstudio.ai

marimo - A Next-Generation Python Notebook: marimo.io

Pydantic: pydantic.dev

Instructor - Complex Schemas & Validation (Python): python.useinstructor.com

Diving into PydanticAI with marimo: youtube.com

Cline - AI Coding Agent: cline.bot

OpenRouter - The Unified Interface For LLMs: openrouter.ai

Leafcloud: leaf.cloud

OpenAI looks for its "Google Chrome" moment with new Atlas web browser: arstechnica.com

Watch this episode on YouTube: youtube.com

Episode #528 deep-dive: talkpython.fm/528

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#528: Python apps with LLM building blocks

Today we’re digging into the Model Context Protocol, or MCP. Think LSP for AI: build a small Python service once and your tools and data show up across editors and agents like VS Code, Claude Code, and more. My guest, Den Delimarsky from Microsoft, helps build this space and will keep us honest about what’s solid versus what's just shiny. We’ll keep it practical: transports that actually work, guardrails you can trust, and a tiny server you could ship this week. By the end, you’ll have a clear mental model and a path to plug Python into the internet of agents.

Episode sponsors

Sentry AI Monitoring, Code TALKPYTHON

NordStellar

Talk Python Courses

Links from the show
Den Delimarsky: den.dev

Agentic AI Programming for Python Course: training.talkpython.fm

Model Context Protocol: modelcontextprotocol.io

Model Context Protocol Specification (2025-03-26): modelcontextprotocol.io

MCP Python Package (PyPI): pypi.org

Awesome MCP Servers (punkpeye) GitHub Repo: github.com

Visual Studio Code Docs: Copilot MCP Servers: code.visualstudio.com

GitHub MCP Server (GitHub repo): github.com

GitHub Blog: Meet the GitHub MCP Registry: github.blog

MultiViewer App: multiviewer.app

GitHub Blog: Spec-driven development with AI (open source toolkit): github.blog

Model Context Protocol Registry (GitHub): github.com

mcp (GitHub organization): github.com

Tailscale: tailscale.com

Watch this episode on YouTube: youtube.com

Episode #527 deep-dive: talkpython.fm/527

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#527: MCP Servers for Python Devs

Today, we’re talking about building real AI products with foundation models. Not toy demos, not vibes. We’ll get into the boring dashboards that save launches, evals that change your mind, and the shift from analyst to AI app builder. Our guide is Hugo Bowne-Anderson, educator, podcaster, and data scientist, who’s been in the trenches from scalable Python to LLM apps. If you care about shipping LLM features without burning the house down, stick around.

Episode sponsors

Posit

NordStellar

Talk Python Courses

Links from the show
Hugo Bowne-Anderson: x.com

Vanishing Gradients Podcast: vanishinggradients.fireside.fm

Fundamentals of Dask: High Performance Data Science Course: training.talkpython.fm

Building LLM Applications for Data Scientists and Software Engineers: maven.com

marimo: a next-generation Python notebook: marimo.io

DevDocs (Offline aggregated docs): devdocs.io

Elgato Stream Deck: elgato.com

Sentry's Seer: talkpython.fm

The End of Programming as We Know It: oreilly.com

LorikeetCX AI Concierge: lorikeetcx.ai

Text to SQL & AI Query Generator: text2sql.ai

Inverse relationship enthusiasm for AI and traditional projects: oreilly.com

Watch this episode on YouTube: youtube.com

Episode #526 deep-dive: talkpython.fm/526

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#526: Building Data Science with Foundation LLM Models

Building a UI in Python usually means choosing between "quick and limited" or "powerful and painful." What if you could write modern, component-based web apps in pure Python and still keep full control? NiceGUI, pronounced "Nice Guy" sits on FastAPI with a Vue/Quasar front end, gives you real components, live updates over websockets, and it’s running in production at Zauberzeug, a German robotic company. On this episode, I’m talking with NiceGUI’s creators, Rodja Trappe and Falko Schindler, about how it works, where it shines, and what’s coming next. With version 3.0 releasing around the same time this episode comes out, we spend the end of the episode celebrating the 3.0 release.

Episode sponsors

Posit

Agntcy

Talk Python Courses

Links from the show
Rodja Trappe: github.com

Falko Schindler: github.com

NiceGUI 3.0.0 release: github.com

Full LLM/Agentic AI docs instructions for NiceGUI: github.com

Zauberzeug: zauberzeug.com

NiceGUI: nicegui.io

NiceGUI GitHub Repository: github.com

NiceGUI Authentication Examples: github.com

NiceGUI v3.0.0rc1 Release: github.com

Valkey: valkey.io

Caddy Web Server: caddyserver.com

JustPy: justpy.io

Tailwind CSS: tailwindcss.com

Quasar ECharts v5 Demo: quasar-echarts-v5.netlify.app

AG Grid: ag-grid.com

Quasar Framework: quasar.dev

NiceGUI Interactive Image Documentation: nicegui.io

NiceGUI 3D Scene Documentation: nicegui.io

Watch this episode on YouTube: youtube.com

Episode #525 deep-dive: talkpython.fm/525

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#525: NiceGUI Goes 3.0

Python in 2025 is different. Threads really are about to run in parallel, installs finish before your coffee cools, and containers are the default. In this episode, we count down 38 things to learn this year: free-threaded CPython, uv for packaging, Docker and Compose, Kubernetes with Tilt, DuckDB and Arrow, PyScript at the edge, plus MCP for sane AI workflows. Expect practical wins and migration paths. No buzzword bingo, just what pays off in real apps. Join me along with Peter Wang and Calvin Hendrix-Parker for a fun, fast-moving conversation.

Episode sponsors

Seer: AI Debugging, Code TALKPYTHON

Agntcy

Talk Python Courses

Links from the show
Calvin Hendryx-Parker: github.com/calvinhp

Peter on BSky: @wang.social

Free-Threaded Wheels: hugovk.github.io

Tilt: tilt.dev

The Five Demons of Python Packaging That Fuel Our ...: youtube.com

Talos Linux: talos.dev

Docker: Accelerated Container Application Development: docker.com

Scaf - Six Feet Up: sixfeetup.com

BeeWare: beeware.org

PyScript: pyscript.net

Cursor: The best way to code with AI: cursor.com

Cline - AI Coding, Open Source and Uncompromised: cline.bot

Watch this episode on YouTube: youtube.com

Episode #524 deep-dive: talkpython.fm/524

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#524: 38 things Python developers should learn in 2025

Python typing got fast enough to feel invisible. Pyrefly is a new, open source type checker and IDE language server from Meta, written in Rust, with a focus on instant feedback and real-world DX. Today, we will dig into what it is, why it exists, and how it plays with the rest of the typing ecosystem. We have Abby Mitchell, Danny Yang, and Kyle Into from Pyrefly here to dive into the project.

Episode sponsors

Sentry Error Monitoring, Code TALKPYTHON

Agntcy

Talk Python Courses

Links from the show
Abby Mitchell: linkedin.com

Danny Yang: linkedin.com

Kyle Into: linkedin.com

Pyrefly: pyrefly.org

Pyrefly Documentation: pyrefly.org

Pyrefly Installation Guide: pyrefly.org

Pyrefly IDE Guide: pyrefly.org

Pyrefly GitHub Repository: github.com

Pyrefly VS Code Extension: marketplace.visualstudio.com

Introducing Pyrefly: A New Type Checker and IDE Experience for Python: engineering.fb.com

Pyrefly on PyPI: pypi.org

InfoQ Coverage: Meta Pyrefly Python Typechecker: infoq.com

Pyrefly Discord Invite: discord.gg

Python Typing Conformance (GitHub): github.com

Typing Conformance Leaderboard (HTML Preview): htmlpreview.github.io

Watch this episode on YouTube: youtube.com

Episode #523 deep-dive: talkpython.fm/523

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#523: Pyrefly: Fast, IDE-friendly typing for Python

Today we’re turning tiny tips into big wins. Khuyen Tran, creator of CodeCut.ai, has shipped hundreds of bite-size Python and data science snippets across four years. We dig into open-source tools you can use right now, cleaner workflows, and why notebooks and scripts don’t have to be enemies. If you want faster insights with fewer yak-shaves, this one’s packed with takeaways you can apply before lunch. Let’s get into it.

Episode sponsors

Sentry Error Monitoring, Code TALKPYTHON

Agntcy

Talk Python Courses

Links from the show
Khuyen Tran (LinkedIn): linkedin.com

Khuyen Tran (GitHub): github.com

CodeCut: codecut.ai

Production-ready Data Science Book (discount code TalkPython): codecut.ai

Why UV Might Be All You Need: codecut.ai

How to Structure a Data Science Project for Readability and Transparency: codecut.ai

Stop Hard-coding: Use Configuration Files Instead: codecut.ai

Simplify Your Python Logging with Loguru: codecut.ai

Git for Data Scientists: Learn Git Through Practical Examples: codecut.ai

Marimo (A Modern Notebook for Reproducible Data Science): codecut.ai

Text Similarity & Fuzzy Matching Guide: codecut.ai

Loguru (Python logging made simple): github.com

Hydra: hydra.cc

Marimo: marimo.io

Quarto: quarto.org

Show Your Work! Book: austinkleon.com

Watch this episode on YouTube: youtube.com

Episode #522 deep-dive: talkpython.fm/522

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#522: Data Sci Tips and Tricks from CodeCut.ai

English is now an API. Our apps read untrusted text; they follow instructions hidden in plain sight, and sometimes they turn that text into action. If you connect a model to tools or let it read documents from the wild, you have created a brand new attack surface. In this episode, we will make that concrete. We will talk about the attacks teams are seeing in 2025, the defenses that actually work, and how to test those defenses the same way we test code. Our guides are Tori Westerhoff and Roman Lutz from Microsoft. They help lead AI red teaming and build PyRIT, a Python framework the Microsoft AI Red Team uses to pressure test real products. By the end of this hour you will know where the biggest risks live, what you can ship this quarter to reduce them, and how PyRIT can turn security from a one time audit into an everyday engineering practice.

Episode sponsors

Sentry AI Monitoring, Code TALKPYTHON

Agntcy

Talk Python Courses

Links from the show
Tori Westerhoff: linkedin.com

Roman Lutz: linkedin.com

PyRIT: aka.ms/pyrit

Microsoft AI Red Team page: learn.microsoft.com

2025 Top 10 Risk & Mitigations for LLMs and Gen AI Apps: genai.owasp.org

AI Red Teaming Agent: learn.microsoft.com

3 takeaways from red teaming 100 generative AI products: microsoft.com

MIT report: 95% of generative AI pilots at companies are failing: fortune.com

A couple of "Little Bobby AI" cartoons

Give me candy: talkpython.fm

Tell me a joke: talkpython.fm

Watch this episode on YouTube: youtube.com

Episode #521 deep-dive: talkpython.fm/521

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#521: Red Teaming LLMs and GenAI with PyRIT

A couple years ago, Charlie Marsh lit a fire under Python tooling with Ruff and then uv. Today he’s back with something on the other side of that coin: pyx.

Pyx isn’t a PyPI replacement. Think server, not just index. It mirrors PyPI, plays fine with pip or uv, and aims to make installs fast and predictable by letting a smart client talk to a smart server. When the client and server understand each other, you get new fast paths, fewer edge cases, and the kind of reliability teams beg for. If Python packaging has felt like friction, this conversation is traction. Let’s get into it.

Episode sponsors

Six Feet Up

Talk Python Courses

Links from the show
Charlie Marsh on Twitter: @charliermarsh

Charlie Marsh on Mastodon: @charliermarsh

Astral Homepage: astral.sh

Pyx Project: astral.sh

Introducing Pyx Blog Post: astral.sh

uv Package on GitHub: github.com

UV Star History Chart: star-history.com

Watch this episode on YouTube: youtube.com

Episode #520 deep-dive: talkpython.fm/520

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#520: pyx - the other side of the uv coin (announcing pyx)

Today on Talk Python: What really happens when your data work outgrows your laptop. Matthew Rocklin, creator of Dask and cofounder of Coiled, and Nat Tabris a staff software engineer at Coiled join me to unpack the messy truth of cloud-scale Python. During the episode we actually spin up a 1,000 core cluster from a notebook, twice! We also discuss picking between pandas and Polars, when GPUs help, and how to avoid surprise bills. Real lessons, real tradeoffs, shared by people who have built this stuff. Stick around.

Episode sponsors

Seer: AI Debugging, Code TALKPYTHON

Talk Python Courses

Links from the show
Matthew Rocklin: @mrocklin

Nat Tabris: tabris.us

Dask: dask.org

Coiled: coiled.io

Watch this episode on YouTube: youtube.com

Episode #519 deep-dive: talkpython.fm/519

Episode transcripts: talkpython.fm

Theme Song: Developer Rap

🥁 Served in a Flask 🎸: talkpython.fm/flasksong

---== Don't be a stranger ==---

YouTube: youtube.com/@talkpython

Bluesky: @talkpython.fm

Mastodon: @talkpython@fosstodon.org

X.com: @talkpython

Michael on Bluesky: @mkennedy.codes

Michael on Mastodon: @mkennedy@fosstodon.org

Michael on X.com: @mkennedy

#519: Data Science Cloud Lessons at Scale

See the full show notes for this episode on the website at talkpython.fm/134

#134: Python in Climate Science

Lyt når som helst, hvor som helst

Other podcasts you might like ...