Knowledge Architect

Scenario
Scenario

IT

United States

Posted on Jun 18, 2026

01 Mission

  • Own the full knowledge system: documentation, support-bot content, in-app copy, the changelog, and the machine-readable layer that AI assistants ingest.
  • Audit and benchmark against world-class references across documentation and the major model providers, and publish a gap map in month one.
  • Ship documentation the day a model or feature ships, as a release gate, so a customer's engineer and our support bot give the same answer on the same day.
  • Design how knowledge is discovered: information architecture, SEO and GEO (generative engine optimization), internal linking, in-app entry points and layered depth.
  • Turn community and competitive intelligence into canonical knowledge, and feed gaps back to product as signal.

02 Scope of ownership

Owns
  • The full knowledge stack: docs, support bot, in-app copy and changelog
  • Canonical per-model guides
  • The machine-readable content layer that AI assistants can ingest
  • Findability: SEO, GEO (generative engine optimization), search analytics and internal linking
Does not own
  • Original creative examples (the Creative Technologist supplies them; you make them canonical)
  • Support-bot infrastructure and training pipeline (Engineering)
  • Day-to-day support tickets (Support)
  • Brand and marketing narrative (Marketing)

03 What we look for

  • A senior documentation, knowledge or developer-experience role at an AI, developer-tools or software company, ideally with a public body of work still regarded as world-class.
  • Native English and a portfolio of layered, structured writing.
  • A systems thinker, comfortable with taxonomy, versioning and content models.
  • Deeply AI-native: you ship your own tooling, such as a documentation linter or changelog generator, and build something live in the interview.

04 Disqualifiers

  • A librarian posture: maintains what exists instead of building the system.
  • Style over substance: beautiful pages that are wrong, late or missing edge cases.
  • Not AI-native in their own work, with no tool they have built to show.
  • Cannot build: "I'll brief an engineer for that" is the wrong answer here.

05 How we hire2-3 weeks

  1. 01
    Intro call
    30 min
  2. 02
    Product session
    60 min
  3. 03
    Build live
    60-90 min
  4. 04
    Written POV
    Async, 48h
  5. 05
    References
    Parallel
  6. 06
    Founder conversation
    45 min