Free courses for every layer of the AI stack.
Six open-source curricula, organized by what you do relative to a model: work alongside it, build a company on it, govern it, build with it, train it, or run it.
56 role tracks and 426 modules, from first job to principal. Every track is a pair of GitHub repos: one to learn from, one to check your work against. All of it is free and MIT-licensed.
Pick a layer to see its tracks.
- AI WorkforceWork alongside the models
- AI StartupTurn it into a company
- AI GovernanceGovern, secure, and assure AI
- AI EngineeringBuild with the models
- ML EngineeringBuild and train the models
- AI InfrastructureRun the platforms models live on
AI Workforce
Work alongside the models
This org covers AI in everyday professional work: working alongside the models. It's for engineers, analysts, product managers, and managers whose job isn't building AI but whose work now runs through it: delegating to coding agents and assistants, briefing them well, verifying what comes back, and building team habits that make AI-assisted work fast and safe.
5 role tracks, new and being written now.
Start with AI-Assisted Software EngineerVisit the AI Workforce site
Practitioner (L20)
Delegating to coding agents, writing specs & context, reviewing AI-written code, testing as verification, repo harnesses
AI-assisted analysis & SQL, checking generated numbers, notebook agents, reproducibility, communicating uncertainty
Product & Delivery (L30)
AI in discovery & specs, prototyping with agents, judging AI output quality, PRDs as context, shipping faster safely
Leadership (L40+)
Redesigning team workflows, review norms for AI-made work, measuring real productivity, skills & career growth
Org-wide AI adoption: tool selection, usage policy, training programs, champions networks, measuring impact
AI Startup
Turn it into a company
This org covers turning it into a company — the operating knowledge technical founders need: how startups work as a system, customer discovery and product-market fit, go-to-market, finance and fundraising, operations and governance, and the endgame. Each curriculum is one self-contained repository, and you learn by building real founder artifacts, not reading case studies.
8 curricula and 74 modules.
How startups work as a system, the dependency graph, stages, founder operating basics, reading the numbers
Customer discovery, business modeling, runway, pre-seed to seed fundraising, SAFEs & cap tables, founder-led sales
Architecture under uncertainty, build vs. buy, first engineering hires, tech debt as a business decision, scaling
Problem validation, PMF signals, positioning, ICP, founder-led sales, pricing & packaging, demand generation
Discovery to PMF, prioritization, outcome roadmaps, discovery/delivery cadence, experimentation & AI evals
Startup accounting, unit economics, three-statement forecasting, cap tables, convertibles, valuation, fundraising
Entity formation, founder legal architecture, employment law, HR ops, equity-comp policy, compensation
Endgame strategy, M&A structures & acquirer motivations, LOIs, diligence & data rooms, secondaries
AI Governance
Govern, secure, and assure AI
This org covers AI governance — governing, securing and assuring AI: inventories and risk assessments, red-teaming and adversarial ML, release assurance and evaluation evidence, ML security, agentic safety, enterprise control architecture, and executive leadership of an AI governance program.
10 role tracks and 108 modules.
Start with AI Governance AnalystVisit the AI Governance site
Analyst & Risk (L15–L25)
AI inventory & intake, risk assessments, fairness review, impact assessments & model cards, regulatory tracking
Risk frameworks, harm modeling, risk registers, LLM red-teaming, adversarial ML, fairness, privacy risk
Assurance, Privacy, Security & Safety (L30–L45)
Data minimization, PII detection & redaction, memorization & extraction risk, consent & DSRs for models, GDPR/state-law reviews
Release assurance, assurance cases, release gates, evaluation evidence pipelines, cross-jurisdiction obligations
ML security, zero trust, secrets, adversarial ML, runtime security, policy as code, supply chain, SecOps
Agent threat modeling, prompt injection & jailbreaks, dangerous-capability evals, excessive-agency containment
- AI AuditorL45
Audit planning vs ISO/IEC 42001, NIST AI RMF & EU AI Act, evidence sampling, control testing, findings, assurance reports
Architect (L50)
AI control libraries, policy as code, multi-jurisdiction reconciliation, AI management systems, risk appetite
Executive (L60–L70)
Running the program, board & audit reporting, regulator engagement, regulatory change, program maturity
AI regulation & risk, model risk management, responsible AI, AI security, audit evidence, board reporting
AI Engineering
Build with the models
This org covers AI engineering — building with the models: LLM application development, retrieval and RAG, agents and multi-agent systems, evaluation, guardrails, harness engineering, and production operations.
11 role tracks and 70 modules.
Start with Agentic AI DeveloperVisit the AI Engineering site
Builder (L20)
LLM fundamentals, prompting & structured output, tool calling, retrieval basics, first agent, shipping
Engineer (L30)
Trace instrumentation, trajectory & tool evals, LLM-as-judge, RAG evals, eval-gated CI/CD, online evals
Agent loop, frameworks, RAG & memory, multi-agent, evals, guardrails, production, harness engineering
AI feature design, production prompting, RAG in product, orchestration, product agents, guardrails
Ingestion & chunking, embeddings, vector DBs & ANN tuning, learning-to-rank, hybrid retrieval, reranking
Customer discovery, integrating into enterprise stacks, evals on customer data, demo→production, field feedback loops
STT/TTS & speech-to-speech, latency & turn-taking, telephony/WebRTC, vision inputs, voice & multimodal evals
Senior & Architect (L40+)
Internal agent platforms: harnesses, tool & MCP registries, sandboxing, permissions, shared evals & tracing, paved roads
Agent systems from an architecture, eval & observability infra, multi-agent at scale, reliability, cost
Hiring & leveling AI roles, planning non-deterministic work, eval-driven delivery, vendor risk, teams that use coding agents
When not to use an agent, orchestration & memory architecture, eval harnesses, security, durable execution
ML Engineering
Build and train the models
This org covers ML engineering — building and training the models: data pipelines, classical and deep learning, fine-tuning and alignment, evaluation, distributed training, and the senior-to-principal work of setting ML architecture and strategy.
9 role tracks and 82 modules.
Start with Machine Learning EngineerVisit the ML Engineering site
Foundation (L20)
ML foundations, data for ML, classical & deep learning, experiment tracking, deployment, drift monitoring
Specialist & Senior (L30–L35)
Data sourcing & licensing, curation & dedup, labeling ops, synthetic data, contamination checks, dataset versioning
SFT data, PEFT (LoRA/QLoRA), preference optimization, safety tuning, distributed fine-tuning, quantization
Benchmarks, LLM eval harnesses, LLM-as-judge, human eval, agent & safety evals, production eval
- NLP EngineerL30
Tokenization, classical NLP, classification, extraction, QA, summarization, machine translation
ML systems architecture, production LLM integration, experimentation at scale, ML SLOs, technical leadership
Distributed training, data pipelines at scale, cluster orchestration, checkpointing, elastic training, throughput
Staff & Principal (L40+)
Multi-team ML architecture, foundation-model training programs, ML platform strategy, capacity & TCO
Org & industry ML strategy, north-star ML architecture, influence without authority, the technical bar
AI Infrastructure
Run the platforms models live on
This org covers AI infrastructure — running the platforms: foundational Python and Kubernetes through distributed training, GPU and LLM infrastructure, MLOps, platform engineering, performance, and enterprise architecture.
13 role tracks and 92 modules.
Start with Junior AI Infrastructure EngineerVisit the AI Infrastructure site
Entry Level (0-2 years)
Python & ML basics, Linux & Docker, Kubernetes intro, cloud platforms, monitoring
Production ML, distributed training, GPU computing, advanced Kubernetes, MLOps, LLM infra, IaC
Intermediate Level (2-4 years)
Platform architecture, multi-tenancy, serving at scale, platform APIs/SDKs, developer experience
CI/CD for ML, model registry, feature stores, experiment tracking, drift detection, A/B testing
Serving engines (vLLM, TensorRT-LLM, SGLang), batching & KV cache, quantization, routing, GPU autoscaling, latency/cost SLOs
GPU utilization, inference latency, training efficiency, cost optimization, profiling
SLOs for model-backed services, LLM & GPU observability, incident response for non-deterministic failures, provider dependency risk, chaos testing
Advanced Level (4-6 years)
Advanced Kubernetes (operators/CRDs), distributed training at scale, CUDA, multi-cloud, SRE
Enterprise ML architecture, multi-cloud/hybrid, security & compliance, FinOps, HA/DR, LLM/RAG platforms
Leadership Level (6-10 years)
Technical strategy & roadmaps, team building, ADRs, incident & performance management
Cross-org architecture alignment, enterprise standards, multi-year roadmaps, executive communication
Principal Level (8-15+ years)
Deep technical expertise, extreme-scale distributed systems, novel infra solutions, mentorship
Company-wide strategy, multi-year roadmaps, technology selection, architecture governance
How the tracks work
Learn
Each learning repo is a sequence of modules: lecture chapters, hands-on exercises, quizzes, and projects. Start with Module 00 and work forward.
Check your work
The paired solutions repo has a reference walkthrough for every exercise. Attempt each one first; the gap between your answer and the reference is where the learning happens.
Built in the open
The content is written with AI assistance and under ongoing human review, so it can contain mistakes. Report anything wrong through the repo's Issues; fixes ship continuously.