← Home 07 / The JD, answered

Not “explore and infer.” A crosswalk.

Each thing the Head of Content & Curriculum, Education job description asks for, mapped to the exact artifact on this site that proves it. Verbatim from the posting, left; where it is answered, right.

● Docent note

A hiring panel should not have to reconstruct the argument. This is it, made explicit. The quotes on the left are verbatim from the live posting (captured 2026-07-11); the links on the right go to the working demonstration. The headline match is the tenth-from-last bullet — “a thesis about what teaching becomes when every learner has an adaptive AI tutor… here to build that” — because this whole site is that thesis, built rather than described.

What the JD asksWhere it is answered
“Set and defend the quality bar for scaled education. Define what ‘great’ means precisely enough that the team can hit it without you in the room.” The Standards →

The AIHC v03 framework: 11 anchors, each written as a rising performance expectation across four bands — “great” made observable and hittable without the author present.

“The most exciting version of this role encodes its own craft into AI-powered systems rather than guarding it.” The Playbook →

The two-timescale architecture: human craft encoded once into the exemplar + rigor rubric, and Claude generates every lesson against them. The craft is in the system, not the person.

“determine where human craft is non-negotiable… Where that line sits, and how it moves as models improve, is yours to figure out.” The Thesis →

The trust line treated as an environmental-design decision, re-derived as models improve — the standards’ ‘moving line’ (X2) and trust-calibration anchor (B3) operationalize it.

“You have a thesis about what teaching becomes when every learner has an adaptive AI tutor, and you’re here to build that, not a better version of the old thing.” This entire site →

This is that thesis, built. A personal AI tutor that teaches on the learner’s real work, one standard at a time — demonstrated end to end, not described.

“increasingly personalized, not templated courseware” · “learning that adapts to each person and scales far past what any team could produce by hand.” The Learner Tracks →

Forty-eight lessons, each generated on a specific learner’s real work — no two alike. Swap the learner and the Do’s and exit tickets change materially.

“every audience (developers, consumers, enterprise admins, the general public)” Four learner tracks →

One fictional learner per named audience — Maya (developer), Devon (enterprise admin), Priya (consumer), Frank (general public) — each a full 12-lesson arc.

“coordinate them as one production engine rather than a chain of handoffs.” The Playbook & Standards →

The generation pipeline and the orchestration anchor (D1): composed workers, gates, and handoffs run as one accountable system, not a relay.

“You’re rigorous about measurement and grounded in learning science. You define what ‘working’ means… instrument for it, and act on what you find.” The Protocol & the ledgers →

An NGSS/CCSS-grounded method; a derivation chain you can audit; and per-learner ledgers where ‘working’ is transfer and mastery-by-standard, not completion.

“you design for durability and fast regeneration rather than perfecting artifacts that models will outdate.” The Protocol & Regeneration →

Versioned, deprecable, re-derivable standards (D2); superseded versions deprecate, never delete. This very site regenerates from source with one build command.

“at a minimum you can read code, you know what good looks like, and you can coach your technical team members.” Maya’s track →

A full Claude Code × AI-agents developer track — subagent contracts, verification, orchestration — taught at the code level.

“Encoding your taste into systems and workflows that produce far beyond what your own hands could.” The Evidence →

The taste is not asserted — it is 1,702 adjudicated receipts distilled into 52 practice families, the encoded craft the standards are built from.

The short version

The role asks for someone who encodes teaching craft into systems. This site is the proof of exactly that.

A thesis, a standard, its evidence, the playbook that generates from it, and forty-eight lessons it produced — built the way the standards say to build, in the founder’s own brand. Start anywhere.