From Foundation Models to Production — a 20-course Coursera specialization that takes you from generative AI fundamentals on AWS through deterministic agents and multi-modal programming to serverless multi-model architectures.
Enroll on Coursera: AI Tooling Specialization
| # | Course | Focus | Companion Repo | Capstone |
|---|---|---|---|---|
| 1 | Generative AI and Foundation Models on AWS | Tokenization, RAG, Bedrock, llama.cpp, SageMaker Canvas | — | Capstone |
| 2 | Intelligent Applications with Amazon Bedrock | Bedrock console, Claude, Dracula pattern, knowledge bases, agents | — | Capstone |
| 3 | Prompt Architecture and NLP on Amazon Bedrock | Token lifecycle, prompt-as-code, chain-of-thought, Ollama bridge | — | Capstone |
| 4 | AI Orchestration: From Local Models to Cloud | Prompt pyramid, caching, Ollama, llamafile, GPU computing, Spot | — | Capstone |
| 5 | Enterprise AIOps and Amazon Q Business | Q Business, CloudShell, cost control, MLOps, RAG workflows | — | Capstone |
| 6 | AI Security and Governance on AWS | Guardrails, CloudTrail, auth patterns, SageMaker Clarify, Rust | — | Capstone |
| 7 | AI-Powered Analytics and Performance Engineering | Lambda, Rust, Amazon Q, CodeCatalyst, benchmarking | — | Capstone |
| 8 | CLI Automation with Amazon Q and CloudShell | Q CLI, Docker, CDK, Lambda, ECR, infrastructure as code | — | Capstone |
| 9 | Deterministic LLM Programming and Quality Metrics | Code quality, AST analysis, technical debt, PMAT, Elo ratings | deterministic-llm-coding | Capstone |
| 10 | Agentic AI: Actor Models and Subagent Architecture | Actix, Rust, Go, Deno, supervision trees, subagents | agentic-ai | Capstone |
| 11 | AI-Assisted Debugging and Test-Driven Fixes | AI debugging, TDD, logging, context gathering, bug discovery | ds500-debug-with-ai | Capstone |
| 12 | Multi-Modal AI: Screenshots to Production Code | Copilot, screenshot-to-code, Playwright, MCP, visual programming | multi-modal-programming-course | Capstone |
| 13 | Privacy-Conscious Development with AI Assistants | GitHub Advanced Security, Dependabot, Grype, secure prompting | windsurf | Capstone |
| 14 | AI-Powered Data Pipelines with Deno | Deno tasks, pre-commit hooks, quality gates, pipeline automation | data-pipelines-deno-typescript-course | Capstone |
| 15 | Building Deterministic MCP Agents | MCP, provable contracts, property testing, fuzz testing, Kani BMC | deterministic-mcp-agents | Capstone |
| 16 | Conversational Bot Architecture with Rust and Deno | Tokio, async runtime, memory safety, Discord, Bedrock | universal-bot | Capstone |
| 17 | AI Code Review Automation with GitHub Actions | Actions, code review, LLM prompting, GitHub Marketplace | pmat-action | Capstone |
| 18 | LLM Security: Vulnerabilities and Defense Patterns | Prompt injection, model theft, information disclosure, plugin security | — | Capstone |
| 19 | Build a Production SaaS Application with AI | API design, Docker, GitHub Pages, test harnesses, MVP | wine-api-saas | Capstone |
| 20 | AI Engineering Capstone: Serverless Multi-Model Systems | Cargo Lambda, Bedrock routing, YAML prompts, production deployment | — | Capstone |
git clone https://github.com/paiml/ai-tooling.git
cd ai-tooling
make checkmake help # Show available commands
make lint # Lint markdown files
make test # Validate course structure (20 courses, capstone sections)
make check # Run lint + test
Each course includes a hands-on capstone project that integrates all modules into a realistic scenario. Completed capstones can be shared on LinkedIn as portfolio projects. See the capstones/ directory.
Each course is ~60 minutes of 3–5 minute videos organized as:
Course → Module → Lesson (3–5 videos) → Key Terms + Reflection
Every module ends with a Critical Thinking Assessment (quiz + role-play practice assignment).
- Noah Gift — Founder, Pragmatic AI Labs · Duke University (Courses 1–10, 13, 15, 20)
- Alfredo Deza — Author and content creator · Python, Rust, DevOps, ML (Courses 11, 12, 14, 16–19)
Course content copyright Pragmatic AI Labs. Code examples are MIT licensed.