complete-ai-engineer-bootcamp-2026

# The Complete AI Engineer Bootcamp 2026 - School of AI **Python → Machine Learning → Deep Learning → LLMs → Embeddings → RAG → Agents → MCP → APIs → Evaluation → Security → Deployment → Enterprise AI** A project-driven, zero-to-production curriculum for becoming a modern AI engineer. > The goal is not to finish with impressive prompts. The goal is to finish with systems you can run, test, evaluate, secure, deploy, and explain. ## What students will build Students complete **16 modules, 64 core tutorials, 16 hands-on labs, 12 portfolio projects, 16 quizzes, a 50-question final exam, and an enterprise capstone**. The portfolio projects are: 1. Python AI Utility CLI 2. Data Quality Analyzer 3. Customer Escalation Predictor 4. PyTorch Neural Classifier 5. LLM Support Ticket Triage Service 6. Semantic Search Engine 7. Citation-First RAG Knowledge Assistant 8. Tool-Using Operations Agent 9. MCP Enterprise Tools Server 10. Secure AI API + Evaluation Harness 11. Production Observability & Release Project 12. Enterprise AI Operations Copilot Capstone ## Why this bootcamp is different - Starts from practical Python and reaches production AI engineering. - Covers both classical machine learning and modern GenAI. - Teaches deep learning with PyTorch. - Builds LLM applications with **provider abstraction**, not vendor lock-in. - Uses a deterministic mock provider so the core path does not require a paid AI API. - Supports optional local models through Ollama. - Goes deep on embeddings, vector search, RAG, citations, reranking, and retrieval evaluation. - Builds controlled agents with tools, planning, memory, approval gates, and audit trails. - Teaches **MCP using the current 2026 protocol direction and Python SDK v2 concepts**. - Includes FastAPI, authentication, multi-tenancy, testing, CI/CD, observability, cost, and enterprise architecture. - Treats evaluation and security as core engineering work. - Ends with a capstone that includes architecture, threat model, eval report, runbook, and demo. ## Student outcomes By the end, a student should be able to: 1. Build clean Python AI projects with reproducible environments and tests. 2. Prepare data and train/evaluate classical ML models. 3. Build neural networks with PyTorch. 4. Explain tokenization, attention, transformers, and LLM generation. 5. Design reliable prompt contracts and structured outputs. 6. Implement embeddings and semantic/hybrid retrieval. 7. Build grounded RAG with citations and evaluation. 8. Build tool-using agents with bounded autonomy. 9. Build and secure MCP servers. 10. Expose AI capabilities through production-style APIs. 11. Create repeatable LLM, RAG, and agent eval suites. 12. Threat-model AI systems and test negative security paths. 13. Deploy and observe AI services. 14. Design enterprise AI for identity, scale, cost, reliability, and governance. 15. Present a credible AI engineering portfolio and capstone. ## Recommended learning paths **12-week intensive:** roughly 12–18 hours per week. **24-week part-time:** roughly 6–9 hours per week. **Self-paced:** one module, one lab, and the associated project at a time. ## Curriculum 1. Python Foundations for AI Engineers 2. Data Engineering Foundations: NumPy, Pandas, SQL, and Validation 3. Machine Learning Engineering 4. Deep Learning with PyTorch 5. Transformers and Large Language Model Foundations 6. Prompting and LLM Application Engineering 7. Embeddings and Vector Search 8. Retrieval-Augmented Generation 9. AI Agents, Tools, Planning, and Memory 10. Model Context Protocol (MCP) for AI Engineers 11. AI APIs and Application Integration with FastAPI 12. AI Evaluation, Testing, and Quality Engineering 13. AI Security and Responsible AI Engineering 14. Deployment, MLOps, and Observability 15. Enterprise AI Architecture, Scalability, and Cost Optimization 16. Capstone, Portfolio, and AI Engineer Career Readiness See [COURSE-SYLLABUS.md](COURSE-SYLLABUS.md) for the detailed lesson map. ## Start here 1. Read [SETUP.md](SETUP.md). 2. Read [STUDENT-GUIDE.md](STUDENT-GUIDE.md). 3. Open `curriculum/module-01-python-foundations/README.md`. 4. Complete all four tutorials. 5. Complete the module lab. 6. Take the quiz before opening the answer key. 7. Build the related portfolio project. 8. Update [PROGRESS-TRACKER.md](PROGRESS-TRACKER.md). ## Technology philosophy The course separates durable engineering concepts from fast-moving model and framework APIs. The examples emphasize: - provider interfaces and adapters - local/offline fallbacks - deterministic tests - explicit validation - least privilege - measurable evaluation - replaceable retrieval/model components - clear trust boundaries - observability and rollback ## 2026 compatibility notes Python 3.14 is a current release family in 2026, but the classroom baseline is Python 3.12 or 3.13 to reduce package-compatibility friction. MCP content is aligned to the **2026-07-28 specification direction** and **Python SDK v2** concepts. Security material should be reviewed against the current **OWASP GenAI LLM Top 10** as the ecosystem evolves. ## Repository map ```text complete-ai-engineer-bootcamp-2026/ ├── curriculum/ ├── labs/ ├── projects/ ├── capstone/ ├── quizzes/ ├── notebooks/ ├── datasets/ ├── src/ai_bootcamp/ ├── tests/ ├── docs/ └── .github/ ``` ## Student and instructor resources - [Learning schedules](docs/LEARNING-SCHEDULES.md) - [Progress tracker](PROGRESS-TRACKER.md) - [Assignment map](docs/ASSIGNMENT-MAP.md) - [Official resources](docs/OFFICIAL-RESOURCES.md) - [Student FAQ](docs/course-marketing/FAQ.md) - [Instructor delivery checklist](docs/instructor/DELIVERY-CHECKLIST.md) - [Classroom activities](docs/instructor/CLASSROOM-ACTIVITIES.md) - [Portfolio project checklist](projects/PROJECT-CHECKLIST.md) - [Project README template](docs/templates/PROJECT-README-TEMPLATE.md) - `docs/cheatsheets/` for quick-reference guides - `notebooks/` for guided practice ## Completion standard A strong completion target is: - 80%+ average on module quizzes - 75%+ on the final exam - at least 8 portfolio projects completed - capstone score of 80/100+ - passing automated tests - a live or recorded capstone demo ## License Code examples and written course content are licensed separately. See `LICENSE-CODE` and `LICENSE-CONTENT`. --- **School of AI — The Complete AI Engineer Bootcamp 2026**