// the find
alexeygrigorev/ai-engineering-field-guide
Research into AI engineering interview assignments, take-home challenges, and hiring practices from 2026
A research-backed guide to AI engineering roles, built from ~5,000 scraped job descriptions and aggregated interview experiences. It covers what the role actually is, what hiring looks like in practice, and how to prepare — with company-specific interview data for 51 companies. Aimed at engineers transitioning into AI roles or actively job hunting.
The data foundation is real: job descriptions, scrape scripts, and raw CSVs are all in the repo, so you can verify claims rather than just trust them. The company-by-company interview breakdowns (51 companies with linked source postings) are more actionable than generic prep advice. Learning paths segmented by prior role (DE, DS, MLE, backend, frontend) are unusually practical — each one gives honest time estimates and identifies which skills transfer. The home assignments section pulls from 100+ real GitHub repos of actual take-home challenges, which is something you can't easily find elsewhere.
The primary language is HTML, meaning the actual content is pre-rendered rather than living in editable Markdown — the source format for the published site isn't obvious from the repo structure, which makes contributing awkward. The job market data is geographically narrow (LA, NY, London, Amsterdam, Berlin, India via builtin.com only), which skews conclusions for anyone outside those markets. The repo has no structured way to track whether company-specific interview data is stale — a 51-company table from 2026 will degrade fast as hiring practices shift. The newsletter/course funnel is present throughout, which doesn't undermine the content but is worth knowing.