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

Core course · LLMs & Agents

Trustworthy Machine Learning: Building and evaluating agentic systems

中文:可信机器学习:构建与评估 Agent 系统

Trustworthy agent systems: build a prototype, then use evaluation to make it reliable.

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

Not offered in the 2026–27 ExploreCourses listing.

3 units Not scheduled in 2026–27 Letter or Credit/No Credit ★★★☆☆
AgentTrustworthyEvalProject

Course positioning

CS 329T is a project-based course on the theme 'prototypes are easy, reliability is hard'. It covers foundation models, prompting, RAG, and agentic architectures, with an evaluation method for each; students finish with an evaluation-driven trustworthy agent application.

What it covers

  1. FMs and prompting: capability boundaries, prompt patterns
  2. RAG: retrieval, citations, hallucination control
  3. Agentic architectures: planning, tools, memory, multi-agent
  4. Evaluation: offline metrics, LLM-as-judge, human evaluation, red teaming
  5. Reliability: failure modes, robustness, monitoring
  6. Team project: build and evaluate an agent application

Prerequisites

CS 229 or an equivalent Python ML course; deep learning (CS 230/CS 231N); familiarity with scikit-learn/Keras.

Assessment & workload

Intro assignments (RAG/agent) plus a team project and an evaluation report. Check the current syllabus for details.

Who it's for

People who want to build agent products, AI evaluation, or trustworthy AI; anyone who can build a demo and wants evaluation methodology.

Study advice

Pair with CS 329Z: 329Z is more full-stack engineering, 329T is more trust + evaluation. Define metrics before writing code.

Related courses

Official course description

This project-based course will introduce students to building and evaluating agentic AI applications powered by foundation models. The overriding theme of the course is that building an initial prototype AI system can often be completed easily but refining a prototype into something that is useful and reliable requires iterative improvement based on clear evaluation metrics. We will cover background in foundation models, prompting, and retrieval-augmented generation (RAG) before introducing full agentic AI architectures. For each architecture, the course will study methods for evaluation. Students will complete introductory homework assignments to become familiar with retrieval-augmented generation (RAG) and agentic AI. Students will then work in pairs or small teams to develop applications using agentic or other approaches and evaluate them by adapting evaluation methods presented in the class.Prerequisites: CS229 or similar introductory Python-based ML class; knowledge of deep learning such as CS230, CS231N; familiarity with ML frameworks in Python (scikit-learn, Keras) assumed.