Healthcare RCM operator. AI automation builder.

I spent the last 10+ years inside revenue cycle. Now I build the AI that fixes it.

Most RCM consultants don't ship code. Most healthcare AI builders have never worked a denial queue. I've done both. A decade running credentialing, denials, AR, and appeals inside multi-practice operations, and I build the HIPAA-grade systems that automate the same work I used to do by hand.

• Reporting automation: 835/EOB ingestion, AR aging dashboards, payer mix analysis, denial meeting decks generated from ETL
• Denial workflows: queue ranking by recoverability, appeal-letter generation (UHC / Cigna / BCBS structured letters), payer-form auto-fill (Aetna, Medicare CMS-20027)
• Credentialing systems: provider lifecycle tracking, CAQH/PECOS/license expiration alerts, payer panel status, audit trails
• AI agents for RCM specialists: AR call prep, denial trend anomaly detection, payer policy retrieval, daily denial queue prioritization
• Practice performance analytics: collection forecasting, no-show prediction, contract underpayment detection, provider scorecards

Python · FastAPI · React · TypeScript · PostgreSQL

Cloud-agnostic. AWS, GCP, or Azure depending on the client's existing stack.

My production work is proprietary. HIPAA-grade systems running inside live RCM operations don't get open-sourced. The receipts live on LinkedIn instead, posted as efficiency case stories with the real before/after numbers.

• LinkedIn: https://www.linkedin.com/in/karthikrcm/
• Email: karthikramesh@gravitashc.com