San Ramon, Calif. – September 22, 2026 -- MediKarma, Inc. reported that its proprietary health score achieved a 90.6% correlation with actual claims costs within an 8 million-patient cohort, as the company positions its JackTM and JillTM AI infrastructure as the clinical backbone for direct-to-patient healthcare offerings.
MediKarma's health score correlates at 90.6% with real claims costs across a high-risk patient cohort
The company drew the 8 million-patient cohort from a de-identified population of 65 million patients, selecting individuals based on a minimum claims threshold and membership in the highest-cost disease categories. MediKarma says the correlation demonstrates the value of building patient engagement on an organized, longitudinal clinical picture spanning multiple providers rather than fragmented records.
A 120-day endocrinology pilot shows half of participating patients improved or maintained A1C levels
In a separate clinical pilot conducted with an outside endocrinology practice, 50% of participating patients either reduced their A1C or maintained a healthy baseline over 120 days, the company said.
CMS formally accepts MediKarma into its Health Technology Ecosystem Diabetes & Obesity track
The Centers for Medicare and Medicaid Services accepted MediKarma's application to join the CMS Health Technology Ecosystem initiative's Diabetes & Obesity track, connecting the company through CMS-aligned interoperability networks. MediKarma said it is already live in production, including in active health plan deployments.
JackTM retrieves multi-provider records while JillTM powers provider-reviewed patient engagement
JackTM retrieves and organizes longitudinal medical records across multiple providers and data sources, according to the company. JillTM uses that longitudinal context to support summarization, patient engagement and provider-reviewed workflows where applicable, with both technologies designed to avoid replacing provider judgment.
"Every company trying to reach patients directly today runs into the same wall: too little of the patient's medical record to build a real profile, or too much of it, unsummarized and out of context, to actually use," said Kris Narayan, CEO of MediKarma. "MediKarma exists to solve that specific problem at scale."
MediKarma's Chief Medical Officer warns incomplete charts carry rising clinical and liability risk
Dr. Miguel Venereo, MediKarma's Chief Medical Officer, said most medical records that appear complete are in practice only a fraction of a patient's real history, a gap he said carries clinical and liability risk as more first-pass judgments shift to AI. "An incomplete chart doesn't announce itself; it just looks like a chart," Venereo said.
MediKarma said the same infrastructure is available to pharmaceutical companies, medical device makers, wearable and sensor companies, telehealth platforms and general-purpose AI platforms extending into healthcare, as well as hospitals, medical groups, clinics and payer organizations using the system directly.