Stanford Launches CS329Z: Engineering AI Agents, Building RAG and Agent Loops from Scratch
dotey · x · 2026-10-28
Stanford's new Fall 2026 course CS329Z: Engineering AI Agents (instructors Diyi Yang, Michael Ryan, John Yang) is now public, aimed at aspiring agent engineers.
- Focus: from monolithic LLMs to compound AI systems (LLMs + retrievers + tools + optimizers), covering the full spectrum from pipelines to autonomous agents.
- Core topics: problem selection and decomposition, component choice, data curation, building evaluations, and design tradeoffs.
- Hands-on: students build core components (RAG, tool use, agent loops) from scratch, then learn how frameworks like DSPy abstract these patterns, with two applied assignments and a quarter-long project.
- Lecture 1 covers agent anatomy (LLM core, reasoning, memory, tools, environment) and challenges of reliability, training, safety, and evaluation.
More from coding & agent
- Zero-AI-at-runtime design generator hits run 8, new aesthetics expected by run 10 — Afinetheorem · 2026-10-28
- Bygga: open-source 3D IKEA assembly guides covering 1,171 products, powered by Together AI — nutlope · 2026-10-28
- Gunning for 60fps on a 200MHz CPU: dev shares Dreamcast optimization challenge — AIandDesign · 2026-10-28
- EnvLunch turns executable environments into free verifiable agent training data — KingsCollegeLondon · 2026-10-28
- Dev building Dreamcast game finds Opus 5.5 has nothing left to suggest on speed optimization — AIandDesign · 2026-10-28
- Is agent recovery an underrated problem? Devs debate checkpointing and idempotency — CuteClass9864 · 2026-10-28