AI engineering careers: reverse-engineered from 2,445 real job ads

What it does

This is a curated research repository that breaks down AI engineering roles, interviews, and market data by analyzing real-world sources. Alexey Grigorev scraped 2,445 job descriptions from cities like LA, NY, London, Amsterdam, Berlin, and India, then synthesized the findings into role definitions, skill expectations, and interview guides. The result is a living document covering everything from system design questions to learning paths for backend engineers wanting to pivot.

The interesting bit

The unusual angle is empirical stubbornness: instead of opining on what AI engineers should know, the author reverse-engineered 5,694+ job responsibilities and 4,525 use cases from actual listings. It even catalogs interview processes at 51 specific companies and mines 100+ GitHub repos for real take-home assignments, treating hiring like a dataset rather than a vibe.

Key highlights

• Grounded in 2,445 scraped job descriptions across six global markets
• Interview questions consolidated from 100+ sources, spanning theory, coding, system design, and behavioral rounds
• Company-by-company breakdowns for 51 organizations, linked to original postings
• Role-specific learning paths (e.g., backend engineers: 2–3 months; data engineers: 3–4 months)
• Raw and structured YAML datasets included for independent analysis

Caveats

• The repo is a work in progress; salary analysis and community-contributed interview experiences are marked as coming soon
• Some sections naturally funnel toward the author’s paid course and newsletter, though the core data remains freely accessible
• Webinars and event recordings are hosted externally, so the repository acts more as a curated index than an interactive tool

Verdict

Care if you are transitioning into AI engineering and want evidence-based prep rather than LinkedIn platitudes. Skip it if you are looking for a hands-on framework or automated tooling—this is a research digest, not a library.