I didn't set out to build a healthcare SaaS company. I set out to stop watching my own accounts receivable age into write-offs.
Crown Valley Imaging is a radiology practice in Mission Viejo I've run as CEO for 15 years. Like every healthcare practice in America in the last decade, we've watched patient-responsibility balances grow from a small slice of AR to a dominant one. High-deductible plans did it. Cost shifts did it. The result is the same: every month we had a six-figure number of patient balances aging out of 30-day collection windows into 60, 90, 120, and eventually write-off.
The vendor search that went nowhere
For a while I did what most healthcare operators do: evaluate vendors. I looked at Collectly, Cedar, Patientco, smaller point solutions. The decks were all fine. The pricing was what you'd expect (1.5–4% of collections plus implementation). The case studies were all “+30% average lift” against unspecified baselines, which, if you've been around healthcare AR for a while, you know means very little.
The bigger issue was that none of them was what my practice actually needed. We needed AI-powered outreach, sure. We also needed settlement campaign management, payment plans with auto-draft, inbound SMS triage, dispute routing, and a send-time optimizer tuned for our specific patient population. Each vendor had some of those features; none had all of them. And the ones with the most features locked you into their payment processor, which meant another hand in the till.
The pitch I kept hearing was: “We'll collect 30% more for you in exchange for 3% of everything we collect.” The math didn't work for me. I knew what 3% of collections was in dollars. It was more than the headcount I was trying not to hire.
The decision to build
Here's the part that sounds crazier in retrospect than it felt at the time: I decided to build it myself. Not “hire developers to build it for me” — actually build it, with my own hands, using Claude Code and Cursor and the stack I'd been teaching myself for two years.
I'm not a developer. I'm a 15-year healthcare CEO who'd gotten comfortable reading code, shipping small internal tools, and stitching together APIs. I'd been operator-at-the-keyboard for enough small wins to believe I could do something big.
60 days from concept to production. The stack was Retell AI for voice, Twilio for SMS, Stripe for payments, Make.com for orchestration, and a ton of custom code to glue the AI layer together. Twelve observable AI capabilities — settlement campaign generation, inbound SMS intent classification, autonomous safe-intent replies, dispute manager, risk scoring, send-time optimizer, a payment-plan negotiator, and a semi-autonomous follow-up agent that load-balances work across the human staff.
We went live December 15, 2025.
What the first 123 days looked like
I'm going to be honest about what happened next because every case study I've ever read in healthcare finance glosses over the numbers. I won't.
Over 123 days of post-launch operation — a window I chose because it matched the final 123 days of the legacy platform before cutover, for apples-to-apples comparison — the platform collected $535,623 in patient-responsibility dollars. The legacy platform had collected $233,471 over its equal-length last window. +129.4% on the same patient panel.
Per paying account: $182.62 vs. $133.87 on legacy, or $160.35 if you strip out everything except the pure digital rail. Both comfortably above HFMA MAP Keys' published benchmark range of $110–$150 for imaging groups of comparable size.
90-day responsibility collection rate: 75.8% vs. 53.1% on the legacy platform, and ~15 points above the industry median. Cost-to-collect: 1.39% — below HFMA's 2–3% best-practice benchmark for digital-first platforms.
Why I'm telling you this
Because a lot of owners in my position — running $2M–$15M businesses, watching vendors charge 3% of collections, feeling like the tools aren't quite right — are one good weekend and some Claude Code subscription money away from building the exact thing they need. Not selling it. Running it in their own business.
Veredge exists because of what this platform did for my practice. After three months of watching the numbers land, I did the thing I didn't set out to do: I started helping other healthcare practices and, because the operator's lens travels, other SMBs in other industries build their equivalent.
The patient-collections platform might eventually get spun out and licensed. I haven't decided. What I know for sure is that I'm going to keep running it on my own ledger, because the math works there — and what's built by the operator actually running the business tends to fit the business better than what's built by a vendor trying to fit 100 practices at once.
If you have a problem, a small technical aptitude, and four weekends, you can build the fix yourself — or you can hire somebody who has.— The operator's takeaway
That's the whole pitch. Book a 20-minute call if you want to talk about what you'd build if you had the time.