AI-Assisted Software Engineering
Master AI-assisted development for Silicon Valley. Build production apps with AI—the right way.
This course trains master's level computer science students to become Silicon Valley-ready software engineers in the age of AI. Students will master AI-assisted development tools while maintaining professional engineering standards, learning to build high-quality, production-ready full-stack applications.
Building successful software with AI requires more than good prompting—it demands understanding how Large Language Models work, when to use which tools, how to evaluate AI-generated code systematically, and how to integrate AI into professional development workflows.
Students will master three distinct AI coding paradigms:
Best for: Architecture planning, learning, complex problem-solving
Skills: Claude Projects, artifacts, conversational iteration
Best for: Professional development, production code, daily workflow
Skills: Tab autocomplete, Composer, .antigravityrules, YOLO mode
Best for: Automation, multi-file refactoring, DevOps
Skills: Terminal workflows, autonomous agents, script automation
All projects are open topic—you choose what to build after getting instructor approval.
Individual project: a personal utility application that solves a real problem validated through Mom Test interviews.
Pair project: a full-stack application built with the IDE harness, Agile practices, TDD, and CI/CD.
Pair project: deploy a production-grade app demonstrating mastery of Claude Code extensibility (W9-W14).
Navigate through the semester to see what you should be working on each week.
Schedule is tentative and subject to change.
| Week | Topics | Slides |
|---|---|---|
| 1 | Introduction | Intro |
| 2 | LLM Architecture & Tokenization | LLMs |
| 3 | Prompt Engineering Basics | Prompts |
| 4 | Claude Web & Artifacts | Research |
| 5 | Claude Web Deep Dive: Artifacts, HW1 due (Mon), HW2 due | Artifacts |
| 6 | Project 1 Due, IDE-Centric AI Coding, P2 pair formation | IDE AI |
| 7 | Agile/Scrum + Pair Workflow | Agile+Pairs |
| 8 | Advanced IDE AI Features, HW3 due | IDE Advanced |
| 9 | Project 2 Due, Claude Code Foundations, P3 pair formation | Claude Code Foundations |
| 10 | Claude Code Workflows & TDD, spec-driven development, HW4 due | Claude Code Workflows |
| 11 | Claude Code Extensibility | Claude Code Extensibility |
| 12 | Agent Architectures & SDK, HW5 due, (Fall Break Nov 25–29) | Agent Architectures |
| 13 | AI Security & Code Quality | AI Security |
| 14 | Production & Course Synthesis | Production |
| Finals | Finals Week — Project 3 Due (Dec 14–20) |
| Component | % |
|---|---|
| Participation | 15% |
| Weekly Quizzes | 10% |
| Homeworks (5 assignments) | 25% |
| Projects (3 major) | 50% |
Organized by week. Click to expand each section.
Slack is the main communication channel for this course. Canvas is used for assignment submissions and grades.
This course requires AI tool use but with strict guidelines: document all AI usage, understand all code submitted, never commit code you cannot explain. Weekly quizzes test conceptual understanding.