AI-Augmented Engineering
Outcome: AI in your toolkit
You’re not behind — you just haven’t pointed your existing skills at AI yet. This path isn’t “learn AI from scratch.” You already know how to read a diff, debug an incident, search a codebase, and write SQL. Every project here takes one of those things you already do well and puts an LLM inside the workflow you already understand — so you’re applying judgment you’ve already built, not starting over.
The 12 projects below move from “AI assists one task” to “AI runs as infrastructure inside your team’s process” — by the capstone, you’re not using an AI tool, you’re shipping one that gates real pull requests.
A CLI that reads a git diff, chunks it per file, and uses an LLM in JSON mode to return prioritized, structured findings by severity and category. You already know what a good code review looks like — this project is about encoding that judgment into a tool, not learning review from scratch.
Proves you can wire an LLM into an existing engineering workflow rather than bolt one on as a side feature — the core skill behind 'AI-Augmented Software Engineer' roles.
What's next
If you want to go broader into ML engineering, owning training, serving, and monitoring on AWS, Path 3 is the natural next stop. If you would rather slow down and cover the foundational LLM patterns this path assumed you already had, Path 1 builds those from scratch. Both are independent entry points, not a required sequence after this one. ML Engineering on AWS →