careers product · sourced 2026-09-27
Get hired as an AI Engineer or AI Architect
What FAANG/MAANG and the frontier labs actually test in 2026, which hands-on
projects get people hired, and a map from where you are on this site to the role you want —
built from the same public postings and interview-loop research every card below cites.
5 roles mapped
sourced & dated
reads your own progress
free, no login
⚠️ Hiring facts change fast. Every figure on these pages carries its
"as reported on" date and its source. Treat them as a starting point for research, not a
guarantee — confirm the loop and the bar with your own recruiter before an interview.
Start here
Five pages are live today. All are interactive: tap through them rather than
reading a wall of text.
🗺️ Role map & readinessAI Engineer, Applied
AI/FDE/Solutions Architect, AI Architect, ML Engineer, Research Engineer — what each is hired
for, its loop, the site path that prepares for it, and a live readiness read from your own
progress on this site.
📊 What top companies want (2026)
The skills 198 AI companies actually ask for, and how Meta, Google, Amazon, Anthropic
and forward-deployed roles test for them — every number sourced and dated.
🔎 JD decoderPaste a job description and get
the skills it asks for (required, responsibility or nice-to-have), the site pages that cover
them, and your gaps against your own progress. Keyword-based, no LLM, nothing uploaded.
🏢 Company loop libraryGoogle, Meta, Amazon,
Microsoft, Apple, Netflix, NVIDIA, Anthropic, OpenAI and Databricks: the rounds, what each
evaluates, the notable format and how to prepare — every claim dated and sourced, with a
confidence note per company.
🧮 GenAI system-design roundThe five-step
framework, a numbers sheet, and twelve worked designs — ChatGPT, RAG, agents, voice and more — each
with tokens/s, KV-cache memory, replicas and cost per 1,000 requests, a 45-minute timer and a
self-grade rubric.
🧭 The PathNot a careers page — the
roadmap itself. The role map's readiness bars are computed from what The Path already knows
you've done.
Coming next
Planned cards, in the order they're queued. Nothing below is live yet — this list
exists so the hub stays honest about what it doesn't have.
- The GenAI system-design round — 12 worked designs with a timed-practice mode.
- An AI-enabled coding round simulator, and 30 LLM-adjacent coding problems.
- A customer-discovery / FDE simulator with take-home briefs.
- Portfolio projects: eval engineering, an LLM gateway, agent reliability, an AI-architect
capstone, and packaging every flagship for a résumé.
- A full mock loop with a behavioral story bank and a readiness dashboard.