Physician-founded healthcare AI
What if the most important clinical finding was the one no one noticed?
Clinical Foresight helps physicians recognize what matters in complex medical records, before the next clinical decision or note is finalized. One platform for clinical safety, prospective Clinical Documentation Integrity, and a path toward autonomous CDI.
Better Clinical Decisions.Better Documentation.Better Patient Care.
The premise
Important clinical information is buried in records that keep getting more complex.
Physicians make high-stakes decisions under real time pressure, across admission notes, histories, medication lists, labs, imaging, and specialist recommendations. Traditional CDI arrives afterward and reviews what was written. Clinical Foresight is built to work before the note is signed.
- Reviews longitudinal clinical information, not isolated documentation or code matching.
- Integrates patient safety and prospective CDI in a single physician-facing workflow.
- Built around evidence-supported recommendations, explainable findings, and physician judgment.
- Patient
- Test, Synthetic A.
- Age
- 71
- Day
- Hospital day 3
- Admit dx
- Community-acquired pneumonia
- Sources
- H&P, 2 progress notes, 38 lab results, MAR, chest imaging
Creatinine 1.1 → 1.6 → 2.0 mg/dL over 48 h. Progress note plan unchanged. Vancomycin trough pending; last dose given 6 h ago.
Ask: urine output trend, contrast exposure, nephrotoxic agents. Consider documenting acute kidney injury if criteria are met.
Apixaban listed on home medications; held on admission. No order and no documented rationale since day 1.
O₂ 4 L NC, RR 26, pCO₂ normal on day-1 gas. Clinical evidence may support a more specific diagnosis if the physician agrees.
Two products. One clinical foundation.
Fix documentation at the point of care, and automate what still needs fixing afterward.
Each product is independently valuable. Together they form a strategy for safer care, fewer preventable documentation gaps, and a progressively automated CDI operation.
Product one · Prospective
AI Hospitalist + Prospective CDI
Clinical intelligence at the point of care. Safety surveillance, decision support, and documentation review that accompany the physician from pre-rounding through note finalization.
- Surfaces overlooked conditions, medication concerns, and unanswered questions
- Suggests high-value questions to investigate at the bedside
- Combined safety and CDI check before every note is signed
Product two · Retrospective
Autonomous Retrospective CDI
From manual chart review to intelligent automation. Screens inpatient records against clinical evidence, identifies documentation gaps, and prepares compliant physician queries.
- Deterministic CMS CC/MCC and MS-DRG reference logic
- Three-phase path from reviewer-assisted to autonomous workflows
- Human oversight concentrated on exceptions
Built around how hospitalists actually work
Three moments in the day. One continuous clinical review.
The review follows the physician through the hospitalization rather than arriving as a query days later.
Review the record, surface what's easy to miss
The AI reads available admission documentation, histories, medications, labs, studies, and specialist notes. It flags potentially overlooked safety concerns and proposes high-value questions for the encounter.
Integrate what the physician learned
The physician adds updated history, exam findings, and decisions. The system folds that into its assessment and raises further questions or concerns where appropriate.
Combined safety and CDI review
A final pass identifies issues that warrant attention and opportunities to document supported diagnoses more precisely. The physician reviews, decides, and signs.
Clinical Foresight is in prototype and integration development. A read-only Epic SMART on FHIR connection has retrieved synthetic sandbox patients; the physician-authored clinical engines are being integrated. The platform is not yet deployed in production hospital environments, and clinical effectiveness has not yet been validated.
How the products work together
Get documentation right the first time. Automate the rest.
Traditional CDI today
Review after the fact
- Charts are read by hand, after documentation is complete.
- Queries reach physicians days later, out of clinical context.
- Safety and documentation are handled by separate teams and tools.
- Cost scales with headcount; consistency depends on the reviewer.
With Clinical Foresight
Prospective clarity, then intelligent automation
- The AI Hospitalist resolves documentation gaps while the physician is treating the patient.
- Retrospective CDI screens what remains and prepares compliant queries.
- One clinical evidence foundation serves both workflows.
- Long-term: retrospective correction becomes increasingly unnecessary.
Why it's different
Technology shaped by the realities of practicing medicine.
Developed from real clinical practice
Grown from a physician-developed AI Hospitalist workflow refined through daily use, not designed around administrative processes.
Longitudinal, not isolated
Reviews the full clinical story across the stay rather than matching codes against single documents.
Safety and CDI in one workflow
Medication, diagnostic, and treatment concerns surface alongside documentation opportunities, in a single physician-facing view.
Standards-based integration
Designed for hospital EHRs, including Epic, through SMART on FHIR interfaces. Read-only by design today.
Explainable findings
Every item cites the clinical evidence behind it, so the physician can evaluate and decide quickly.
Progressive autonomy
A deliberate path from reviewer-assisted CDI to automated queries, coding support, and downstream revenue-cycle preparation.
Built from the bedside outward
Clinically consequential details and documentation opportunities get missed, even by experienced clinicians, when the information is dispersed across the record. Bring another layer of clinical intelligence into the moment the physician reviews the patient and finalizes the note.
Health systems · Technology partners · Investors
Let's talk about what your organization could do with prospective CDI.
Initial conversations and demonstrations use synthetic data. Any real-patient pilot requires separate institutional authorization, security review, and clinical evaluation.
DrFrancis@clinicalforesight.ai