Winter Quarter 2026

Building & Scaling LLM Applications

CS 224G is Stanford's premier project-based course for building production-ready AI applications. Master reasoning models, agentic workflows, and responsible AI development with industry leaders.

10 Weeks
4 Sprints
60 Students
∞ Possibilities
View Schedule Watch Demo Day 2025
future_of_ai.py 
from llm import ReasoningModel
from agents import MultiAgentSystem
from rag import ContextEngine

class CS224GProject:
    """Your next AI startup begins here."""
    
    def __init__(self):
        self.model = ReasoningModel("o3")
        self.agents = MultiAgentSystem()
        self.context = ContextEngine()
        
    async def build_the_future(self):
        # 10 weeks to launch
        return await self.ship_it()|

Schedule

Tue & Thu

10:30am – 11:50am

Location

In Person

420-040

Credits

Letter Grade

3-4 Units

Team Size

Project Teams

2-5 Students

About the Course

The LLM Age Demands New Builders

Conversation has become code. The model is the CPU. A new universe of application development has opened up—and there is far more bad information than good.

CS 224G is a project-based course designed to explore the frontier of LLM applications. You'll learn their capabilities and limitations, then apply cutting-edge techniques to program them efficiently and effectively. Form teams, prototype apps, and iterate in a startup-style environment with two-week sprints culminating in Demo Day.

This is not just another AI course—it's managerial training. You're about to scale yourself with an AI team. We'll help you get it into shape while building something innovative.

What You'll Master

🧠Reasoning Models (o1, o3, R1)
🤖Agentic Workflows & Multi-Agent Systems
📚Context Engineering & RAG 2.0
⚡Production Deployment & Observability
🔊Multimodal AI & Voice Systems
🛡️AI Safety, Ethics & Guardrails
📊Evaluation & Red Teaming
🚀AI Entrepreneurship & Pitching
Meet Your Instructors

Learn From Those Who've Built

John Whaley

John Whaley

Instructor

CS PhD, Serial Entrepreneur. Exits from UnifyID + Redcoat. Founder of Inception Studio.

LinkedIn
Jan Jannink

Jan Jannink

Instructor

CS PhD, Serial Entrepreneur. Exit from VoiceBase. Founder of Synthpop AI.

LinkedIn

Course Assistants

Andrea Mock
Andrea Mock
amock@stanford.edu LinkedIn
Ahmed Sharaf
Ahmed Sharaf
ammas@stanford.edu LinkedIn

Office Hours

Jan, Andrea & Ahmed Wed 2:00pm - 4:00pm Gates 358
John Thu 11:50am - 1:00pm 420-040
Prerequisites

What You Need

Required

Python Proficiency

CS 106B or equivalent programming experience

Required

ML Fundamentals

CS 221, CS 229, or equivalent understanding

Required

ML Framework Experience

PyTorch, TensorFlow, or JAX

Recommended

API & LLM Experience

Familiarity with OpenAI, Anthropic, or similar APIs

💡

Don't Have Formal Prerequisites?

Students with substantial practical experience building with LLMs may enroll with instructor permission. Email cs224g-win2526-staff@lists.stanford.edu with your background and interest.

Evaluation

How You're Graded

100%
10% Participation Class attendance & Slack engagement
45% Sprint Checkpoints 3 bi-weekly demos & code reviews
45% Demo Day & Final Technical excellence, product thinking, presentation
Logistics

Everything You Need to Know

📍 Lectures

Tuesday & Thursday 10:30am-11:50am in person at 420-040. Attendance is mandatory.

🎥 Recordings

The class is not recorded or available via Zoom. Be there live to get the full experience.

💬 Communication

All course-related questions on Slack. For personal matters: cs224g-win2526-staff@lists.stanford.edu

♿ Accommodations

Initiate requests with the Office of Accessible Education (OAE). Send accommodation letters to the staff email.

Ready to Build the Future?

The class fills quickly. Join the Slack, submit your project proposal, and secure your spot in CS 224G Winter 2026.

Join Course Slack View Project Guidelines