Tracking my journey into AI engineering, one module at a time. Fundamentals to agents, with hands-on projects. Open to all.
A structured, opinionated curriculum for software engineers moving into AI/ML engineering — the discipline that separates someone who can call an LLM API from someone who can design, evaluate, ship, and operate AI systems that hold up in production. It's built as plain markdown: seven modules from foundations to frontier, each with canonical resources, the tools actually used in industry, a hands-on project, and the pitfalls that bite people first. I'm writing it as I learn (publicly, mistakes included), and it's open for anyone making the same transition — self-taught devs, backend/full-stack engineers adding AI to their scope, and new grads who want depth beyond "I used the OpenAI SDK once."
Last reviewed: August 2026 · See hot-topics.md for what's moving fast vs. what's settled.
Module What you'll be able to do -------------------------------------- 01 LLM Fundamentals Explain what the model is actually doing, and pick one on evidence 02 Prompting & Context Engineering Get reliable, structured output and manage the context window as a budget 03 Retrieval & RAG Build retrieval that measurably finds the right thing 04 Agents & Tool Use Design agent loops that terminate, recover, and stay in budget 05 Evaluation & Observability Replace vibes with numbers, offline and in production 06 Deployment & AI Infra Serve models at a known cost, latency, and failure profile 07 Emerging Topics Evaluate new capabilities without chasing demos
From the project README.
Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:
[](https://mcp.liqiwa.com/s/bimal1023--AI-native-engineer.html)
Give the agent "this" and he gets a real PC.
metahub-ai/assayAn open, reproducible trust layer for AI artifacts — skills, MCP servers, agents, and plugins. Assay inspects code, runs sandboxed behavioral audits, and generates verifiable reports. Help us build th
junxchan54-maker/mcd-cn-assistant🍟 麦当劳中国官方 MCP 智能助手技能：把 35 个原子工具编排成 4 条业务流 —— 活动推送 · 联名关注 · 最优惠点餐 · 多人群点聚合账单。算价用 calculate-price 真算，下单必二次确认。｜McDonald's China MCP agent skill: campaigns, collab watch, best-value ordering, group bill s
graphersal/graphersalIn-memory graph database and Gremlin traversal engine in Rust: browser playground, CLI, MCP server for AI agents, durable store, Python bindings.
wahyunuriman999/Muse-Skill-HubOpen-source, LLM-agnostic capability runtime for AI agents — 97 skills as typed, permission-enforced MCP tools. Final certification: 23/23 gates PASS.
JUSICK/Argos-mcp-guardrailA lightweight, sub-millisecond security shim that prevents AI agents from accessing sensitive files (.env, SSH keys) and executing destructive commands via MCP.
The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.