Courses that end with something you shipped
Two families of course. The Claude Series is a set of short, hands-on masterclasses on a single Claude tool, open to anyone. The bootcamps are project-based programs for teens building a portfolio.
Which course should I take?
- Want to get hands-on with one Claude tool quickly — pick a masterclass from The Claude Series. No experience needed, and each one is self-contained.
- A teen building a coding portfolio — go to the bootcamps. New to coding? Course 1. Already writes Python? Course 2.
- Not sure between Course 1 and Course 2 — a quick readiness check: can you write a function that uses
if,for, andreturn, and run a script from the terminal? If yes, you're ready for Course 2. If not, Course 1 first. - Already comfortable with Python and calling an LLM API, and want proof of a real contribution — Course 3, our advanced track.
Series · 3 masterclasses
The Claude Series
Short, hands-on masterclasses, each focused on a single Claude tool. Built for total beginners — students, professionals, freelancers, and founders — and every one includes hands-on labs, so you leave with a real deliverable rather than notes.
Claude Code: The Terminal Architect
Build & Ship Real Apps — Without Writing a Single Line of Code
- Live online
- 8 segments · 2 live builds
- No experience needed
Direct Claude Code in plain English: set the project brain, describe the outcome, review and correct, then ship and automate. You leave with a deployed app you keep and one automation already running on a timer.
- A working app built and deployed live, with no code written by hand
- CLAUDE.md as your project brain — the single highest-leverage habit
- Routines and /schedule, so tasks keep running after class ends
- MCP connectors for Gmail, Calendar, and web search
Course details
Claude Design: The Visual Architect
From Prompts to Production Workflows
- Online masterclass
- 2 modules
- Beginner to practical builder
Use Claude Design as a creative partner instead of a blank canvas — describe what you need, refine it with real context, and turn it into a clickable prototype that is backed by editable code.
- Chat-and-canvas: every design lives as code you can edit element by element
- Import a brand system so colors, type, and components stay consistent
- Build interactive prototypes and design live with your team
- Hand off to Claude Code with /design-sync, or publish to a live domain
Course details
Claude Cowork: The Executive Orchestrator
From Desktop Setup to Automated Workflows
- 2 hours
- Online
- 2 modules + 2 labs
- Beginner to practical user
Turn Claude from a chat assistant into a desktop coworker that organizes workspaces, creates real files, connects your apps, and runs scheduled work — with permissions and approval loops throughout.
- Safe setup: sandboxes, permissions, and approval loops
- Projects and folder instructions that make Claude follow your rules
- Real files out: documents, spreadsheets, PDFs, structured reports
- Connectors (Google Drive, Slack, Zoom MCP), screen control, mobile dispatch
Course details
Bootcamps · 3 courses
Project bootcamps for teens
Three courses, one path. Course 1 takes a complete beginner to their first shipped app. Course 2 takes someone who already writes Python into building real AI agents. Course 3 takes it further: join a real working AI project and ship one checkable contribution for your portfolio. Course 1 and 2 run 5 weeks; Course 3 is self-paced over 4. Every course ends with something you can open, run, and show.
Course 1 · Beginner class
AI Coding — Learn Python with Cursor
- 5 weeks
- Online
- No experience needed
Starting from zero, students learn Python inside Cursor — using AI to generate, edit, and debug code — and turn an idea into a real app they ship and present.
What you'll learn
- Week 1 — Python basics — variables and data types, writing and running real code in Cursor from day one
- Week 2 — Data structures, functions, and loops — organizing scattered code into logic that solves a problem
- Week 3 — OOP, deeper Cursor workflows (code generation, editing, debugging, agent mode), and GitHub version control
- Week 4 — APIs, JSON, and the web — pulling in real data and building the final project
- Week 5 — Demo Day — polish, record a demo, and present live
What you'll finish with
- One original app you built and can run — a small game (Tetris, 2048) or a useful tool
- A public GitHub repo holding every weekly project
- A recorded demo walkthrough, presented live on Demo Day
- A completion certificate, on finishing the course and meeting the grade bar
- Prerequisites:
- None. Built for absolute beginners — no coding experience required.
- Format:
- Each week: 2 recorded lessons (watch anytime) + one 2-hour live session (teaching and Q&A). Bookable 1:1 TA sessions and a regular office hour throughout. 9 lectures across 4 teaching weeks, then a Demo Day week.
Explore Course 1
Course 2 · Advanced class
AI Agent — From Chatbot to Autonomous Agent
- 5 weeks
- Online
- Basic Python required
Five weeks, five projects. Students build one new agent capability each week — talking, remembering, acting, planning — then combine all four into the WikiMind capstone.
What you'll learn
- Week 1 · Talks — Persona Chatbot — tokens, prompts, system prompts, and deploying to a public link
- Week 2 · Remembers — Knowledge Agent — context engineering, memory, and retrieval-augmented generation (RAG), so it answers from your sources instead of guessing
- Week 3 · Acts — Research Agent — tool and function calling plus web search, so it gathers and reports back
- Week 4 · Plans — Planning Agent — multi-step planning, the agent loop, and safety guardrails
- Week 5 · Delivers — WikiMind capstone — one agent that chats, remembers, searches, and plans
What you'll finish with
- 5 GitHub projects — 3 of them live at public URLs anyone can open and try
- The WikiMind capstone as the centerpiece of the portfolio
- The ability to explain LLMs, system prompts, RAG, tool calling, and the agent loop in a sentence each
- A completion certificate (PDF), on finishing the course and meeting the grade bar
- Prerequisites:
- No AI background needed, but basic Python is required: you can read and write a function containing if, for, and return, and run a script from the terminal. Starting from zero? Take Course 1 first.
- Format:
- About 4 hours a week — a 2-hour live lecture plus a 2-hour TA-led workshop — roughly 20 hours across the 5 weeks.
Explore Course 2
Course 3 · Advanced class
Forward Deployed AI Engineer — From Working System to Your Own Contribution
- 4 weeks
- Online
- Python + LLM APIs required
Students join a working AI reference project, finish one real, checkable task inside it, and leave with proof of their own contribution that they can explain and defend in interviews.
What you'll learn
- Week 1 · Understand — Reference project onboarding. Read a running AI system, run it locally, and map how its pieces fit together.
- Week 2 · Contribute — One standard task sized to your target role, such as an evaluation lab: design a rubric, assess real outputs, and iterate on prompts.
- Week 3 · Review — One in-depth review and a revision round, then turn the work into resume bullets and a portfolio case study.
- Week 4 · Present — An interview story and live mock interviews, so you can walk through what you built and why.
What you'll finish with
- Evidence of what you actually built: your own code diffs, test records, or a reproducible notebook, depending on your target role
- The engineering judgment to explain it: inputs and outputs, architecture, failure cases, trade-offs, and which parts were yours
- A career-ready story: role-relevant resume bullets, a GitHub or portfolio case study, LinkedIn Projects content, and interview coaching
- Verification of your training and project completion (a training record, not proof of an internship)
- Prerequisites:
- You're comfortable with Python and calling an LLM API.
- Format:
- Self-paced over 4 weeks. Materials come as a GitHub learning package, with on-demand support from Azure Partners mentors.