Med Foresight was started because the people building it had spent two decades watching electronic health records take time away from patients — not as observers, but as the physician in the room.
Med Foresight was started by two co-founders — one who practiced medicine for 22 years, and one who has spent his career building large-scale data and AI systems for institutions that depend on getting them right.
22 years practicing gastroenterology. Every clinical decision in the product passes through her.
Washington University in St. LouisGastroenterology fellowship
Leads the clinical work — what surfaces in a chart summary, how the AI handles diagnostic uncertainty, when the product should defer to physician judgment.
A career building large-scale AI systems for institutions that cannot afford to get them wrong.
Leads the AI architecture — the proprietary medical language model, its validation, and the integration with hospital systems.
A few principles shape how Med Foresight is developed, validated, and deployed. They are not slogans. They determine what we ship and what we will not.
Med Foresight produces drafts, summaries, and answers. The treating physician reviews and approves every output before it enters the patient record. The AI does not write into the chart unsupervised — the physician carries clinical responsibility.
We work with a pilot GI practice before we expand. We measure what happens — physician time saved, summary accuracy, override rates — and we are honest about where the product defers.
Med Foresight is built to support HIPAA-compliant workflows and works with HL7 and FHIR interfaces. The product is designed to be reviewed by a compliance team, not to surprise one.
The chart summary and physician chatbot are live today. They are the first two capabilities of a system being built across the full GI visit cycle — preparation, consultation, follow-up.
US physicians spend approximately half of their working time inside electronic health records.1 At national scale, that represents roughly $200 billion in lost productivity each year. Behind that number is something simpler: time that should have gone to patients.
The vision Med Foresight works toward is not new. The medical community has discussed it for more than a decade — a record that anticipates clinical need rather than burying it.2 The question is no longer whether such a system should exist. The question is whether it can be built with the care that medicine requires.
Med Foresight does not aim to solve healthcare. It aims to give physicians back the parts of their day that should never have been taken from them.
Med Foresight is built to support HIPAA-compliant workflows and works with HL7 and FHIR interfaces.
