method

the lanes

the corpus was produced by parallel agent research lanes — one on company history and the people, one on coverage, sentiment, and the independent-lab idea — each writing a lane report against a shared scratch source catalog before synthesis. the pattern is the company-research formula proven on roam-research and the other dossiers on this site: lane-prefixed source ids frozen upfront, per-claim confidence, and a declared deduplication owner. synthesis merged the lane catalogs, remapped aliases to canonical ids, and cut claims no surviving source could carry.

the evidence rules

1. every event, quote, and claim carries at least one source id.
2. dates keep their precision — a year, a month, or a day — and approximate dates are labeled rather than sharpened.
3. confidence stays honest: confirmed, reported, and inferred are distinct claims, not stages of the same claim.
4. sources are typed by role — primary, interview, reporting, archive, community — so a claim's foundation is visible before it is read.
5. sentiment labels describe the coverage, not the truth of it.

known gaps

1. ownership — no cap table, equity split, or cap-table-adjacent document has ever been disclosed; the self-funded claim rests on holz's own statements and is not independently verified.
2. subsidiaries — none are publicly evidenced; the corporate perimeter beyond midjourney, inc. is a finding, not a footnote.
3. the founding date — holz's linkedin says august 2021; a san francisco business registry lists a start date of 2020-09-16. the discrepancy is preserved, not reconciled.
4. the org chart — current titles beyond a handful of confirmed roles come from aggregators (rocketreach, everydev) and are marked reported, not verified.
5. revenue, headcount, and user figures — all self-reported, press-sourced, or litigant-alleged; the $500m–$600m claims rest on weak aggregators; ~21m discord members are not paying users.
6. the api — no public api exists and no official statement of intent was found; every enterprise complaint is inference from absence.
7. niji × spellbrush — the collaboration's terms, revenue share, and corporate structure are undisclosed.
8. hardware continuity — whether the 'orb' 3d-capture program survived as a track separate from the medical scanner is unconfirmed.
9. office hours — no official transcripts exist; the record is community recaps of varying reliability, and quotes carried through them are marked accordingly.
10. the exact date holz left ultraleap — 'august 2021' is the linkedin figure; no departure filing or contemporaneous account fixes it tighter.

corrections

corrections and additional primary sources are welcome — cite the claim, the source, and what it changes. the catalog is designed to be audited, not just read.