- Contact
[withheld] — Machine-restructured corpus of the entire US insurance regulatory filing archive ([withheld]): 25 years of rate filings, actuarial memoranda, Excel raters and rule manuals converted from PDF/spreadsheet into structured, queryable data and into executable rating models in a documented JSON schema, with source links back to the originating filing page. Maximum data depth at minimum company size: a 25-year, 25TB, 1.2M-filing national corpus distilled into executable rating models, built by a 5-person seed-stage firm with ~$1M funding. That asymmetry — corpus far exceeding company…