About AI Engineer Path

This is the resource I wish existed when I went from "I can ship a REST API" to "I just put an AI feature in front of real users." A year of building production AI at work, six months of writing it down clearly, plus every reader who pushed back on something and made an article better.

Who this is for

Backend engineers learning to build with AI. Concretely:

• You can ship a REST API, query a database, run a container.
• You have never called an LLM API and you do not know what tokens are.
• You want to add AI features at work, not write academic papers.
• You are short on time and impatient with fluff.

If that is you, every article on this site is written for you. If you are an ML researcher or just learning to program, you will be better served elsewhere.

What you will find here

• 13 articles across foundations, RAG, evals, and production patterns, plus a capstone that wires it all into a working service.
• Real code (Python, real libraries, runnable) instead of pseudocode.
• Diagrams when they help, prose when they do not.
• Honest "this is the hard part" notes on the things that are actually hard, and equally honest "skip this for now" notes on the things that are not.
• A small glossary of the terms used across articles, each linked to the piece that introduces it.

What you will not find: hype, "10 prompts to revolutionize your workflow," or content written by an AI dressed up as a human.

How it is run

This site is a side project. Written by one engineer (me), edited by no one, fact-checked against production work that shipped. There is no team, no content calendar, no quarterly OKRs. Articles ship when they are ready, typically a few per month.

If something on the site is wrong, send a note via the contact page. It gets fixed within a day. That is the entire support model and it works because the inbox is just me.

Why it is free

Charging would gate the people who would benefit most: backend engineers at smaller companies, students, engineers in regions where a $99 USD course is real money. The donation model (one-time tip, monthly supporter, annual supporter) keeps the lights on without putting content behind a wall.

If the site helps you ship something at work, consider chipping in. If not, that is fine. Read everything for free, and recommend it to a friend who would get value from it.

The voice

Practical, not academic. Opinionated but honest about trade-offs. Direct, not corporate. Short sentences. No em-dashes (they are the single biggest AI-generated tell). Light on metaphors because the reader is technical and does not need "imagine your AI is a librarian."

If a sentence reads like it could have been generated by an LLM, it gets rewritten until it does not.

Where it goes from here

The current curriculum (AI Engineering, 13 articles) is the first complete topic. MCP Development is the next planned topic; the first article is live, the rest are coming.

Topic ideas after that include agents and tool use, LLM fine-tuning, multi-modal models, and voice / real-time AI. The order depends on what readers ask about most. If you have a suggestion, the contact page reaches me.

Get in touch

Best place: the contact page. Replies usually come within a day.

Last updated 2026-05-26.