8 weeks · 3 sessions a week · about 3 hours each

The curriculum

Every week: foundations, going deeper, then a build lab where you demo and explain what you built.

  1. Week

    01

    AI-native workflow & product thinking

    How professionals build with AI, and how to pick a problem worth solving.

    1. S1 · FoundationsHow software is built with AI today. Tool setup, Git, and your first AI-built feature.
    2. S2 · Going deeperWriting specs and context the AI can follow. Reviewing AI changes line by line.
    3. S3 · Build labTeams form, interview users, and scope a version 1 that fits 8 weeks.

    You ship: Product brief, screen sketches, and a first reviewed commit

  2. Week

    02

    Web frontend

    Pages, components, and forms in Next.js — and how to spot bad AI-generated UI.

    1. S4 · FoundationsHow a web app is put together: routes, components, props, and state.
    2. S5 · Going deeperPrompting for clean, responsive, accessible UI. Fixing what the AI gets wrong.
    3. S6 · Build labBuild the product's main screens with sample data.

    You ship: All core screens working on desktop and phone

  3. Week

    03

    Backend, database & auth

    Real data, real accounts, and the security work you never leave to AI.

    1. S7 · FoundationsData modeling, tables, relationships, and how a request reaches the database.
    2. S8 · Going deeperSign-in, row-level security, and validation. Testing that users only see their own data.
    3. S9 · Build labConnect every screen to real data with full create, edit, and delete.

    You ship: Working sign-in and data that stays private per user

  4. Week

    04

    AI features in the product

    Calling language models from your backend, reliably and affordably.

    1. S10 · FoundationsHow LLMs work: tokens, context, system prompts, and keeping API keys on the server.
    2. S11 · Going deeperStructured JSON output, validation, streaming, failures, and cost per user.
    3. S12 · Build labShip one genuinely useful AI feature inside the product.

    You ship: One AI feature live, with failures handled

  5. Week

    05

    RAG & vector databases

    Make the AI answer from your product's own data, with sources.

    1. S13 · FoundationsEmbeddings and vector search, and why RAG beats fine-tuning for most apps.
    2. S14 · Going deeperChunking, pgvector, retrieval quality, citing sources, and per-user privacy.
    3. S15 · Build labBuild “chat with your data” and test it against real questions.

    You ship: AI answers grounded in the app's own data

  6. Week

    06

    AI agents

    Let the AI take real actions — safely.

    1. S16 · FoundationsWhat an agent is: tool calling and the agent loop.
    2. S17 · Going deeperPermissions, confirm-before-action, guardrails, and prompt injection.
    3. S18 · Build labBuild an agent that performs one real action, and test it with tricky inputs.

    You ship: An agent that safely takes a real action

  7. Week

    07

    Mobile & deployment

    A Flutter app on the same backend, and everything live in production.

    1. S19 · FoundationsFlutter with AI: screens, navigation, and calling the shared backend.
    2. S20 · Going deeperDeploying to production: hosting, environment variables, secrets, and an Android build.
    3. S21 · Build labShip the mobile core flow and gather feedback from real users.

    You ship: A live web app and an installable Android app

  8. Week

    08

    Quality & demo day

    Testing, explaining your system, and presenting it to an audience.

    1. S22 · FoundationsTesting and debugging with AI, and evaluating AI behavior.
    2. S23 · Going deeperExplaining your architecture and pitching your product in three minutes.
    3. S24 · Build labDemo day: every team presents, and every student explains their part live.

    You ship: A live demo, a documented repository, and your individual walkthrough

Ready to build?

Not covered: training models, advanced data science, multi-agent systems.

Reserve your seat