12-05, 18:30–19:00 (UTC), AI/ML Track
In this talk we present the OS library Burr -- a tool that makes it easier to build reliable, production-ready AI applications and agents. We will show how to use Burr to address a host of production concerns problems including generating test data from prior runs, interactive debugging, persisting/loading application state, and more.
Building AI applications is hard. While an abundance of tooling makes it easy to get a flashy demo out in front of your manager, shipping an app to production (and real customers) requires solving a host of complex challenges -- avoiding hallucinations, integrating observability, managing persistence, and plugging back into the rest of the business, among more.
In this talk, we'll present Burr -- a highly customizable open source library that addresses all these challenges easier through a simple function-based python API. Specifically, we will show how Burr can help you:
- Monitor the decisions your application makes at every point
- Add persistence/telemetry with little additional effort
- Load up a production app at any point in time and debug
- Handle complex human-in-the-loop feedback
- Gather data for future test-cases/eval
We will go over the basic concepts of the library, walk through a simple example, demonstrate how you can leverage the above features easily, and present the ecosystem of plugins/integrations and future roadmap for Burr.
No previous knowledge expected
Elijah has always enjoyed working at the intersection of math and engineering. More recently, he has focused his career on building tools to make data scientists and researchers more productive. At Two Sigma, he built infrastructure to help quantitative researchers efficiently turn ideas into production trading models. At Stitch Fix he ran the Model Lifecycle team — a team that focuses on streamlining the experience for data scientists to create and ship machine learning models. He is now the CTO at DAGWorks, which aims to solve the problem of building reliable AI systems through open source software. In his spare time, he enjoys geeking out about fractals, poring over antique maps, and playing jazz piano.