Introducing Pydantic's new platform
summary
Announcement shows a concrete observability platform: configure with import and one line to start sending traces, get nested spans that record start/end times, attributes and exceptions, and automatic DB and HTTP instrumentation that finds the slow SQL that made a real endpoint jump from 46ms to 1s. It extracts attributes from f-strings, offers autotracing wraps every function, OpenTelemetry under the hood, Pydantic model pretty views, SQL querying of traces, LLM/OpenAI panels, and live/demo-ready demos.
AI-written summary · about
chapters (15)
- 0:00Why Logfire and what problem it solves
- 6:42Linear logs to structured traces
- 11:17Python-first observability
- 15:39Auto-tracing and function insights
- 17:10Pretty Python objects in logs
- 18:39Database query tracing
- 22:27Web framework integrations
- 24:02HTTP request tracing
- 26:47Distributed tracing
- 27:55Other language support
- 29:53Pydantic integration insights
- 33:03LLM and AI tracing
- 35:42Custom UI panels
- 37:11SQL querying in Logfire
- 39:19Q&A
related talks
17mEfficient Python Observability with OpenTelemetry
24m"Introduction to OpenTelemetry with Python"
26mIntro to Pydantic, run-time type checking for your dataclasses
59mLightning Talks
32mWhat was your Python up to at 3am?
57mLightning Talks - Part 2
28mDeveloping Culture To Write Reliable And Performant Services At Scale
30mPython 3.12's new monitoring and debugging API