A Chrome extension that does five jobs at once: writes your clinical charts, catches undercoding in real-time, learns how your clinic bills, bridges your EHR + clearinghouse + payer portals, and turns your accumulated clinic data into insights you can act on. One sidebar. Zero tab-switching. No API, no IT, no setup.
Built for independent clinics and multi-specialty groups who can't afford enterprise revenue cycle management — and shouldn't have to.
| Specialty | Avg Revenue Lost to Undercoding | Cheryl Impact | Source |
|---|---|---|---|
| Family Medicine / Internal Medicine | $10,000-50,000/yr | E/M upgrade capture (99213→99214, ~$30/visit) | AAFP, AAPC audit data |
| Dermatology | $20,000-75,000/yr | Excision size coding, layered closure billing, modifier optimization | Industry case studies |
| Orthopedics | $30,000-100,000/yr | Modifier errors (59, RT/LT, 22), implant coding, unbundling | ASC facility-level data |
| Psychiatry / Psychology | $10,000-50,000/yr | 90834→90837 upgrade ($46-56/session), add-on code 90785 | Multiple billing sources |
| Urgent Care / Walk-in | $15,000-60,000/yr | E/M level optimization, G2211 add-on, POS code correction | Per-visit extrapolation |
| Multi-Specialty Groups | 3-8% of revenue | Cross-specialty coding consistency, centralized audit | Industry standard (MGMA) |
| Cardiology | $25,000-100,000/yr | ECG interpretation codes (93005/93010), stress test components | Cardiology billing data |
| Physical Therapy / OT | $8,000-25,000/yr | 8-minute rule optimization, timed vs untimed code accuracy | PT industry data |
| Chiropractic / Acupuncture / TCM | $5,000-20,000/yr | Modifier -25, -AT (Medicare), E/M + procedure same-day | ACA data |
Sources: AAFP (2018), AAPC audit data (2021), PMC10026023 (Florida Medicare study), ACA (2024), PT Everywhere. Ranges are per provider/year estimates. MGMA DataDive has proprietary specialty benchmarks.
| Component | Detail |
|---|---|
| Frontend | Chrome Extension (Manifest V3), sidebar panel |
| AI Engine | GPT-4o (OpenAI) with function-calling for structured outputs |
| Billing Engine | Rule-based: ICD-10-CM/PCS validation, NCCI edits (Q1 2026), MPFS fee schedule (CY 2026), MDM scoring matrix per 2021 E/M guidelines |
| Data | 30+ specialty billing rule sets, CMS-sourced coding rules, updated quarterly |
| Integration | Browser-based screen reading — works with 40+ web-based EHRs (see compatibility list), zero API dependency. Desktop-only EHRs are not supported. |
| Security | HIPAA-compliant. BAA signed with OpenAI for API data processing. No persistent patient data storage on our servers. All data encrypted in transit (TLS 1.2+). Extension reads only visible screen content. |
1. Provider opens their EHR in Chrome and clicks "Read Page" in the Cheryl sidebar.
2. Cheryl reads the visible screen content (patient info, chart history, intake forms) — the same data the provider already sees. (Riding along in clearinghouse and payer portals is on the Phase 1 roadmap.)
3. Provider types shorthand notes (e.g., "bp 120/80, a1c 7.2, cont metformin, f/u 3mo") or asks Cheryl in natural language.
4. Cheryl generates a complete chart and automatically runs the 17-step billing pipeline — catching missed codes, suggesting upgrades, flagging NCCI conflicts, and linking each suggestion back to the exact chart sentence that justifies it.
5. Provider reviews, accepts suggestions, and pastes back to EHR. Every correction trains Cheryl's per-clinic memory — the third repeat becomes a permanent rule for the whole team.
6. Behind the scenes, Cheryl silently captures deidentified operational metadata (codes, timing, denials, payer mix). The private clinic dashboard that surfaces these insights is on the roadmap — the capture layer is already in place so early members’ data builds the baseline. No PHI ever leaves the EHR.
First 50 providers: $49/mo. Less than the cost of one upgraded visit.
Get Early Access →