Tools

Built from real client work, not a whiteboard.

Each one started as a solution to a specific client problem, then got generalized. They run as their own product, separate from the consulting practice.

Live

Health Center Database

Every U.S. federally qualified health center — operations, financials, contractors and audit findings — assembled from the primary federal filings rather than from a summary of them. Roughly half of these centers withhold their financials from HRSA, so the IRS Form 990 is not a second opinion on them; it is the only revenue figure that exists, and the database says which is which.

  • 1,512 active Section 330 grantees and look-alikes, with seven years of UDS behind each
  • Contractor spend parsed out of Form 990 Part VII-B, resolved to one name per vendor across every spelling filed
  • Every center's own reported EHR — platform, product and version — plus who switched, from what, and when
  • Material weaknesses and the corrective action plan text, from the Federal Audit Clearinghouse

Open healthcenterdb.com →

Live

List Cleaner

Conference and prospect lists routinely arrive with names that don't match your CRM cleanly — typos, rebrands, generic org names. List Cleaner matches an uploaded list against your CRM using domain-first identification, the same signal priority a careful human reviewer would use.

  • Confirmed matches, a human review queue, and net-new candidates — never a silent auto-merge
  • Domain exact match, domain variant, redirect/rebrand detection, then phone and name signals, in that order
  • Company and contact matching run independently, then cross-referenced — catches job changes and new contacts at known organizations
Live

Dedup & Audit

Separate from list matching, Dedup & Audit scans an entire connected CRM in one pass — companies and contacts together — looking for the duplicates that accumulate over years of manual entry, imports, and merges nobody tracked. Every candidate pair gets an evidence-tier score. Nothing merges until you open the review screen and confirm it yourself.

  • Scans the whole portal, not just an uploaded list — built for the audit you run once a duplicate problem is suspected
  • Companies and contacts scanned together from one shared pass
  • Every pair scored with evidence tiers; nothing merges without an explicit per-pair or bulk confirmation
  • Deep-audit and ongoing-watch modes for after the initial cleanup
New

Card Capture

Photograph the cards from a conference booth — business cards, attendee badges, handwritten "enter to win" slips — and Card Capture turns the photos into verified leads. A vision model reads every field, flags handwriting it isn't confident about instead of guessing, and hands the result to the same matching engine List Cleaner uses.

  • Ambiguous handwriting is held for human review, never silently guessed — a misread field can't overwrite a real CRM value
  • Zoom re-read on uncertain fields, using the original photo resolution, not a downscaled copy
  • QR and badge codes decoded as identity proof, not treated as contact data
  • Staff annotations on cards (pricing notes, "no contact," prize winners) captured verbatim

Questions before you try it on your own portal?

Email me directly