LLM Engineering in Action
The book I’m writing about the LLM stack from useful foundations to systems that need to run, fail, recover, and be debugged.
AI engineer
Mohammad Shojaei
I build LLM systems and write about the weird parts: agents that lose state, retrieval that sounds right until it isn't, evals that replace vibes, and model-serving bills that quietly become a plot twist.
This page is only the index. The longer biography and work history live on /about.
Start here
LLM Development Roadmap
A practical path through LLMs. Not a sacred curriculum. More like a map with coffee stains and warning labels.
Building Production AI
Short field notes on agents, retrieval, evals, serving, debugging, and the boring glue that makes AI systems survive users.
Elsewhere
GitHub
Tools, experiments, and small engineering artifacts. Some are useful. Some are suspiciously alive.
Medium
Longer writeups when a bug, paper, or late-night experiment turns into something worth explaining.
Hugging Face
Models, datasets, adapters, and evaluation leftovers from the build loop.