A self-driving, open-source AI bootcamp powered by a local LLM tutor -- learn by building 15 real projects, zero cloud API keys required.
12 phases | 35 lessons | 15 projects | ~28 hours | 100% local
Most AI tutorials teach theory or require expensive cloud API keys. This bootcamp is different:
- 100% local -- everything runs on your machine via Ollama, no API keys or cloud accounts needed
- Learn by building -- 35 hands-on exercises + 15 real projects, not just reading
- AI-graded -- every exercise is graded by real tests (60%) + AI rubric (40%)
- Socratic AI tutor -- a local LLM guides you with questions, not answers
- Progressive -- starts from "What is a terminal?" and ends with deploying full-stack AI apps
- Progress tracking -- submissions, quiz scores, and lesson completion saved automatically. Works without an account; sign in to sync across devices or for team tracking
┌─────────────────────────────────────────────────────────┐
│ learner's machine │
│ │
│ ┌───────────┐ REST ┌──────────────┐ Ollama │
│ │ Next.js │ <──────> │ FastAPI │ <──────> │
│ │ Web UI │ │ Tutor Engine │ llama3.2 │
│ │ :3000 │ │ :8000 │ :11434 │
│ └─────┬─────┘ └──────┬───────┘ │
│ │ │ │
│ │ ┌───────────────┘ │
│ v v │
│ ┌──────────────────┐ │
│ │ Supabase │ ┌─────────────────┐ │
│ │ (local Docker) │ │ ACE Framework │ │
│ │ :54321 │ │ generate / │ │
│ │ Auth · DB · │ │ curate / │ │
│ │ Storage │ │ reflect │ │
│ └──────────────────┘ └─────────────────┘ │
└─────────────────────────────────────────────────────────┘
git clone https://github.com/grey8-io/learn-ai-with-grey8.git
cd learn-ai-with-grey8
bash scripts/setup.sh # One-time setup: installs all dependencies, pulls AI model
bash scripts/start.sh # Starts all services (Ollama, Tutor API, Web UI)Open http://localhost:3000 to start learning.
For prerequisites and detailed instructions, see the Getting Started guide.
Know a developer who wants to learn AI? Send them this repo. No signup, no API keys, no subscriptions — they can start building in 5 minutes.
If you're a team lead or engineering manager, this is a free upskilling path for your developers. Share it in your team Slack, forward it to a colleague, or drop it in your engineering newsletter. The entire course runs locally — no corporate approvals needed.
Quick share link:
https://github.com/grey8-io/learn-ai-with-grey8
| Phase | Focus | Lessons | What You Build |
|---|---|---|---|
| 1 | Dev Environment Setup | 3 | Terminal & Git Basics, Python Environment Setup, VS Code & AI Tools Setup |
| 2 | Python Foundations | 3 | Variables, Types & Functions, Data Structures, Working with Files & APIs |
| 3 | AI Fundamentals | 3 | What is AI? Machine Learning Basics, How LLMs Work, AI Limitations & Ethics |
| 4 | Your First AI App | 3 | Calling LLM APIs, Building a CLI AI Assistant, Milestone Project: AI Chatbot |
| 5 | Prompt Engineering | 3 | Role Prompting & System Messages, Structured Outputs, Few-Shot Patterns & Chain of Thought |
| 6 | Embeddings & Semantic Search | 3 | What Are Embeddings?, Vector Databases, Building Semantic Search |
| 7 | RAG Systems | 4 | RAG Architecture, Document Chunking & Processing, Building a RAG Pipeline, Milestone Project: RAG Knowledge Base |
| 8 | AI Agents | 3 | Agent Concepts & Patterns, Tool Use & Function Calling, Multi-Agent Systems |
| 9 | AI Backend Development | 4 | FastAPI Basics, Async AI Services, Production Patterns, Milestone Project: AI-Powered API |
| 10 | Optimization & Cost Control | 3 | Token & Cost Management, Caching & Batching, Latency Optimization |
| 11 | Deployment & Monitoring | 4 | Docker Basics for AI Apps, Cloud Deployment, Monitoring & Observability, Milestone Project: Streaming Chat App |
| 12 | Capstone Projects | 3 | Project Planning & Architecture, Building & Testing AI Applications, Deployment & Showcase |
Browse the full course guide locally or on GitHub:
| Document | What It Covers |
|---|---|
| Course Guide | Complete curriculum, all 35 lessons, 15 projects, learning cycle, library dependencies |
| Getting Started | Prerequisites, step-by-step setup, health checks, environment variables |
| Contributing | Code style, adding content, PR process, project structure |
Browse documentation:
npx docsify-cli serve docs --port 4000 # Locally at http://localhost:4000Or visit the hosted docs site (once GitHub Pages is enabled -- see below).
GitHub Pages serves the Docsify site directly from the docs/ folder -- no build step, no CI needed.
- Go to your repo on GitHub
- Navigate to Settings > Pages
- Under Source, select Deploy from a branch
- Set branch to
mainand folder to/docs - Click Save
Your docs site will be live at https://<your-username>.github.io/learn-ai-with-grey8/ within a few minutes. Every push to main updates the site automatically.
# Start services (run each in a separate terminal)
ollama serve # LLM runtime
cd tutor && uvicorn tutor.main:app --reload --port 8000 # Tutor API
cd platform/web && npm run dev # Web frontend
# Or start everything via Docker
docker compose -f docker/docker-compose.yml up
# Testing
cd tutor && pytest # Tutor tests
cd platform/web && npm test # Web tests
cd ace && pytest # ACE tests
# Linting
cd tutor && ruff check . # Python lint
cd platform/web && npm run lint # TypeScript lint
# Documentation
npx docsify-cli serve docs --port 4000 # Serve docs locallyTip: If you have
makeinstalled (standard on macOS/Linux; on Windows:choco install make), you can use shortcuts likemake dev,make test,make lint. Runmake helpfor all targets.
- Star this repo to bookmark it and help others discover it
- Follow Nagaraju Mudunuri on LinkedIn for AI career tips and course updates
This bootcamp is built and maintained by Grey8, an AI talent and training company. The curriculum, code, and tutor are free and open-source under AGPL-3.0 — and will stay that way.
Grey8 has plans to offer paid services for learners and teams who want them: proctored certification, candidate placement with hiring partners, enterprise cohort training, and AI consulting. None will be required to use this bootcamp — the free experience is complete on its own.
We welcome contributions of all kinds -- curriculum content, platform features, bug fixes, and translations.
Please read docs/contributing.md before opening a PR.
This project is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0).
You are free to use, modify, and distribute this software. If you run a modified version as a network service, you must make your source code available under the same license. The "Grey8" name and branding are trademarks and may not be used in derivative works without permission.