Stanford launches CS329Z, a full course on engineering AI agents
stanfordnlp · x · 2026-09-25
Stanford NLP has released the first lecture materials for CS329Z: Engineering AI Agents, a new course covering the engineering spectrum from simple LLM pipelines to compound AI systems to autonomous agents.
- Instructors: Diyi Yang (Stanford NLP professor), Michael Ryan (DSPy core contributor), John Yang (creator of SWE-agent/SWE-bench)
- Three core challenges structure the course: decomposition, data, and evaluation
- Modules include LLM building blocks (structured outputs, context engineering, RAG, tool calling/MCP), frameworks like DSPy, agent loops, and eval construction
- Students build core components (RAG, tool use, agent loops) from scratch before learning framework abstractions, culminating in a quarter-long project.
More from coding & agent
- Nemotron 3 Speaker Diarization Ported to Apple Silicon via Core ML and MLX — ivan_digital · 2026-09-25
- Parallel's parallel web search now built into LangChain managed agents — BraceSproul · 2026-09-25
- Same model, 5x cost gap: harness choice matters more than success rate — JeremyCMorgan · 2026-09-25
- TypeSafe AI's Jev evaluation model goes free on Vercel AI Gateway — JohnPhamous · 2026-09-25
- Do web frameworks still matter in the age of vibe coding? A former framework maintainer weighs in — rseroter · 2026-09-25
- Dev runs Opus 5.5 vs Astra cross-audit workflow, Opus concedes 5 times in one afternoon — Ice2jc · 2026-09-25