Towards Useful, Reliable, and Scalable AI Agents
summary
Open-source JavaScript/TypeScript stack called B agent framework plus B code interpreter runs agents with sandboxed Python on Kubernetes, uses Llama-family models (Llama 3.1 and option to run 7B locally), and enforces schemas via Zod-based LLM drivers. Demoed agent calls two tools (web search + Python) to compute US 2030 population and returns a calculable result; observability uses an Observer API and B Observe that logs to MLflow. Serialization, timeouts, max iterations, and guard-rail hooks handle reliability and safety.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:03Defining LLMs and chat LLMs
- 5:40Pitfalls of large language models
- 8:16What is an LLM agent
- 11:07LLM agent challenges
- 15:24Safety, security, privacy
- 18:40B open-source stack
- 21:56B agent framework deep dive
- 25:09Building agent tools
- 26:57Standalone LLMs
- 28:04Structured outputs with drivers
- 29:50Observability in framework
- 31:37Serialization and state
- 32:43Roadmap and features
- 33:26Q&A
related talks
39mTowards Useful, Reliable, and Scalable AI Agents
41mThree agents, three frameworks, one talk
30mBuilding Reliable AI Agents: An Open Source Approach To Evaluation and Observability
40mAgents in Action: Advancing Open Source AI Missions
01h 42mBuilding Responsible AI Agents in Decentralized Open-Source Environments
16mBuilding Agentic AI applications with Positron and AWS Strands Agents
01h 06mBuilding Smart Agents with LlamaIndex: A Comprehensive Overview
19mBuilding Agents at Cloud Scale