🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Why can AIs code for 1h but not 10h? A simple explanation: if there's a 10% chance of error per 10min step (say), the success rate is: 1h: 53% 4h: 8% 10h: 0.002% Toby Ord has tested this 'constant error rate' theory and shown it's a good fit for the data chance of

Watchfiles v1.1.0 released with support for Python 3.14 & pypy 3.11 as well as upgrading the version of the notify crate used. github.com/samuelcolvin/w…

I'm pretty shocked by this from X - I report indisputable phishing, and get a "nothing wrong here" response. WTF are they doing?

x.com/pydantic/statu… The fact that pydantic is cutting new versions Friday at 11PM PT should tell you what you need to know about the team. getting. after. it.

We must be doing something right. OpenAI cohere and IBM are advertising against pydantic ai. 💥

Extremely useful repo: github.com/sirosen/textho…

I've added links to all our competitors that document how to integrate with Pydantic AI. I wonder if any other AI framework + Observability companies are confident enough in their product to do this? 👀 github.com/pydantic/pydan…

PydanticAI version 0.3.3 is out! 🎉 github.com/pydantic/pydan…

A little experiment. I want to trade knowledge. Engineering, AI usage, team management! Book a 30-min call, walk me through your company & the pains you’re facing, and I’ll share my experience. Book👉 cal.com/mitsuhiko/expe…

Me trying to be a dev rel, I think it's safe to say I'm better at writing the code...

I'm chatting to Hugo Bowne-Anderson tomorrow about Human Seeded Evals and Pydantic Logfire on Youtube at 4pm UK. I'm pretty confident something like the think I'm describing here is where we'll end up with evals. Please join, I'd love to hear what you think. lu.ma/voeyx7n1

This morning we hosted a discussion on the MCP ecosystem with operators/founders from StackOne, Pydantic , ElevenLabs, ACI.dev, and more. We covered a lot of ground, from security and auth to reliability when using multiple tools - some of the key takeaways: - We need

Going live with Samuel Colvin in 10 minutes to discuss the current state of evaluation in AI-powered software and what we can do to make it more robust....and human-centered! 🤖 Come say hi 📺 youtube.com/live/wk6rPZ6qJ…

The Pydantic logfire backend, nothing is perfect, but I'm pretty convinced we have the right fundamental stack here. My guess is, if you removed the auto-generated TS from the counts, it would be pretty close to 1/3 .ts, 1/3 .py, 1/3 rs.

We shipped a ton of stuff in Q2 🚀 • Rebuilt Logfire's dashboard engine from scratch: way faster with new Standard Dashboards • Logfire MCP Server: ask your AI "any patterns in recent 404s?" in Cursor and watch it query your live logs & traces • PydanticAI at ~700k

Today’s my last day at Amazon! 😢 Grateful for the journey, but even more for the incredible people I got to work with❤️ Next stop: Polar! I’m joining as the 5th team member of an exciting, fast-growing company! Can’t wait to build what’s next. Stay tuned 🚀

Summer is too hot. Staying inside. Spinning up Modal Installing Pydantic ai Building MCP server

This is great, well worth a look.