BAML - A new programming language for building Agents
summary
BAML is a domain-specific language that embeds inside any host language and turns LLM requests into type-safe LM calls with generated streaming and non-streaming schemas, editor live prompt preview, and enforced test cases. The tool auto-corrects model output to match types (an error-correction system that took eight months to build), powers pipelines parsing 100+ page bank statements with 99%+ documents at $0 cents error and automates health-fax workflows at ~98.5% accuracy, and ships as open source used by 1,000+ companies.
AI-written summary · about
chapters (6)
- 0:00The state of AI pipelines
- 2:42Paradigm shifts in software
- 6:27The hybrid nature of AI code
- 9:33Modeling LLMs and agents
- 16:10Introducing BAML
- 29:12Q&A and conclusion
related talks
19mBAML: a new programming language for AI
01h 28mBuilding LLM Agents Made Simple a
39mTowards Useful, Reliable, and Scalable AI Agents
27mAI Demystified Unveiling the Intelligence Behind the Machine
36mProduction-Ready AI Agents: From LLMs to Small Language Models
01h 43mBuilding LLM Agents (Hackathon)
30mBuilding Robust, Secure LLM and Agentic AI Workflows
41mBuilding LLM Agents – Challenges and Learnings