AI Engineering Track

A structured, 6-stage path through AI engineering — the same shape a focused bootcamp or course would use. Every link below goes to a real, existing, in-depth page on this site; nothing here is a placeholder. If a stage looks thin, that's an honest signal of where this site's own coverage is lighter, not a hidden page.

This is a curated sequence through content that already exists elsewhere on the site (same pattern as Suggested Paths, just organized as a linear course instead of a priority-ranked list) — not a duplicate curriculum.

01 · AI Engineering Foundations

The core skills every production AI engineer needs before touching an LLM: systems, Python, and math.

• CS Fundamentals for AI Engineers — 7 pages
• Python Engineering for AI — 5 pages
• Mathematics for AI — 6 pages
• LLMs & GenAI Overview — the on-ramp into everything below
• Calling the LLM API in Practice — streaming, retries, parameters, and what to log on every call

02 · Building RAG Systems

Retrieval-augmented generation end to end — chunking, embeddings, hybrid search, reranking, and real evaluation, not a toy vector-DB demo.

• Retrieval-Augmented Generation (RAG) — chunking strategies, dense/sparse/hybrid retrieval, advanced RAG architectures, GraphRAG, real eval metrics
• Retrieval & Reranking Architectures — bi-encoder, cross-encoder, late interaction
• Prompt Engineering — structured output, system prompts
• Evaluation & Serving

03 · Agentic AI & Orchestration

Agents that plan, use tools, keep memory, and coordinate — the architecture patterns, not just a framework tutorial.

• Agents — Roadmap — 19 pages: agent fundamentals and architectures, ReAct in production, plan-and-execute, routing & supervisor, reflection/self-critique, context engineering, single-agent vs. multi-agent, multi-agent systems, loops and graphs
• MCP (Model Context Protocol) + MCP Protocol Deep Dive — the open standard for tool/context connectivity
• A2A (Agent-to-Agent)
• Agentic Coding Assistants, Computer-Use & GUI Agents, Browser Automation Agents

04 · LLMOps & Production

Deploying, monitoring, and continuously improving LLM products once they leave the notebook.

• MLOps — Roadmap — 31 pages: containers, Kubernetes, CI/CD & ML CI/CD, model serving, LLM inference engines & optimization, observability, data engineering & versioning, deployment strategies, cloud deployment

05 · Evaluation, Safety & Guardrails

Measuring quality, catching regressions, and holding up under real (and adversarial) usage — the part teams treat as an afterthought until it breaks in production.

• AI Evaluation — Roadmap — 5 pages: evaluation fundamentals, LLM/RAG/agent evaluation, human & adversarial evaluation
• AI Safety & Alignment — Roadmap — 4 pages: alignment & RLHF, scalable oversight & frontier safety
• AI Security — Roadmap — 5 pages: OWASP LLM Top 10 & prompt attacks, AI red teaming & adversarial testing, model & data attacks — the practical guardrails side

06 · AI System Design & Interviews

Architecture tradeoffs and interview-ready reasoning for AI-native systems.

• ML System Design — Roadmap — 6 pages, including real case studies
• Interview Prep — Overview — 9 pages: knowledge Q&A strategy, the quick-reference drill

Want the same ground covered but ranked by how often real job postings actually ask for it, instead of in stage order? See Suggested Paths. Want the full dependency graph across every section on the site? See Learning Path.