Rebrief™ Ambient AI for Surgical and Dental Clinics

What I did

I started Rebrief to solve a problem I was living every day as an OMFS intern: we were spending more time typing than treating. I designed and built an ambient documentation system that sits quietly in the operatory, captures the full encounter, and converts it into:

• Structured progress notes and letters

• Procedure-specific templates for OMFS and dentistry

• Visual odontograms and treatment plans

• Plain-language after-care summaries for patients

From there, I led the end-to-end validation work:

• Clinical design: Scoped workflows with surgeons, residents, and staff at McGill University Health Centre, then adapted them for Harvard and NUS clinics.

• Technical architecture: Prototyped the stack (speech recognition + LLM pipelines) and designed safety rails for hallucination control and PHI handling.

• Research + IRB: Authored multi-site protocols, sample size calculations, and data-sharing plans; navigated ethics and IT approvals across three health systems.

• Product & implementation: Ran pilots in real clinics, gathered feedback on accuracy, time savings, and cognitive load, and iterated the UI/UX to match real chairside constraints.

Results

• Rebrief consistently produced structured drafts of notes and patient instructions that clinicians could review and sign instead of writing from scratch.

• Early pilots showed meaningful reductions in charting time and improved completeness of documentation for complex OMFS cases.

• The project evolved into a multi-institutional research collaboration with co-investigators at McGill, Harvard, and NUS, positioning Rebrief as a real hospital-grade system rather than a demo.

• Rebrief is now the core product of Stat Labs, with IP filed around proactive documentation and predictive workflow automation.