Stanford AI Course Map 中文 GitHub ↗
← Back to the course map
Stanford · 2026—27 CS 329Z LLMs & Agents ★★★★☆

Core course · LLMs & Agents

Engineering AI Agents

中文:AI Agent 工程

Engineering methodology from LLM pipelines to compound AI systems to autonomous agents: choose the problem, decompose it, pick components, and evaluate.

Draft translation: official facts are checked; wording may change after native review.

3 units Autumn 2026 Letter or Credit/No Credit ★★★★☆
AgentEngineeringRAGEval

Course positioning

CS 329Z is one of the closest courses to the AI engineer role, taught by Diyi Yang and others. It emphasizes that an agent is not one model but a compound system of LLMs, retrievers, tools, and optimizers; students build RAG, tool use, and an agent loop from scratch, then move to frameworks such as DSPy.

What it covers

  1. Compound AI systems: design patterns for multi-component systems
  2. RAG: retrieval, reranking, context management, evaluation
  3. Tool use: function calling, tool selection, error handling
  4. Agent loop: planning, execution, reflection, termination conditions
  5. DSPy: declarative optimization of prompts and pipelines
  6. Data curation and evaluation: building eval sets, offline/online metrics
  7. Design trade-offs: latency, cost, quality, controllability
  8. Three applied homework assignments + a quarter-long project

Prerequisites

Programming ability; LLM basics (prompting, API use); ML background is a plus.

Assessment & workload

Three fully applied homework assignments plus a quarter-long project, focused on building, optimizing, and evaluating. Check the current syllabus for details.

Who it's for

Aspiring AI engineers/agent engineers; people moving from backend/ML to LLM applications; founders.

Study advice

It complements CS 224V: 224V is research/formal, 329Z is engineering/eval. Don't just get the demo running — make the evaluation solid.

Related courses

Official course description

The shift from monolithic language models to compound AI systems - systems with multiple interacting components including LLMs, retrievers, tools, and optimizers -n represents a fundamental change in how AI applications are built for people. This course teaches students how to engineer agentic systems: the full spectrum from simple LLM pipelines to compound AI systems to autonomous agents. Students will learn to pick what types of problems to focus on, decompose problems, select appropriate components, collect and curate data, build evaluations, and reason about the design tradeoffs that arise when building these systems in practice. Students first build core components (RAG, tool use, agent loops) from scratch, then learn how frameworks like DSPy abstract these patterns. Through three fully applied homework assignments and a quarter-long project, students will gain hands-on experience building, optimizing, and evaluating agentic systems.