About
I'm Hao Hoang. I trained as a mathematician before I trained models, and the habit stuck: I care less about the answer than about why the obvious answer breaks.
AI Interview Prep started in November 2025 as one interview question at a time. It's now 372 issues across 13 series, read by 23,000 engineers. Each issue puts you in an interview, quotes the question, names the answer most candidates give, and walks through the mechanism that makes it wrong. The format is the point. In a senior interview, half right costs you the offer.
On LinkedIn, where about 74.1k engineers follow along, I write about papers and tools the week they ship, plus the occasional joke about the tools we're all drowning in. The long versions live in two books, LLM System Interview and The RAG Interview.
AI Engineer Labs LLC is the company behind the books and the consulting. If you're building something that works in the demo and not in production, that's the part I like.
Background
- Now
- Senior AI engineer at sourceCode, building generative AI applications for FWD Insurance. Since May 2025.
- Before
- AI researcher and software engineer at Spartan, 2023 to 2025: production LLM systems and agent pipelines. Before that, data scientist at Viettel Big Data Analytics Center, where I built the TV360 recommender end to end, and at Kratos Defense and Security Solutions in France.
- Studied
- Master's in computational and applied mathematics, INSA Toulouse. PhD in AI at the University of Strasbourg, since October 2025.
- Mathematics
- Silver medal at Vietnam's National Mathematics Competition, 2017. A 2022 preprint on how fast inertial dynamics with Hessian-driven damping converge: the continuous-time view of Nesterov's accelerated gradient.
Email: hao.hoang@aiengineerlabs.us