MediKarma Validates “Agentic AI” in Clinical Intervention Study: Demonstrates 50% A1C Reduction; Proves Friction-Free Alternative to ChatGPT

A breakthrough intervention with Dallas Endocrinology proves that autonomous AI agents can bridge the “between-visit” care gap – delivering outcomes that secured a 20,000-unit deployment with Family Health Choice (FHC).

SAN FRANCISCO, Feb. 17, 2026 /PRNewswire/ — MediKarma, an AI-driven platform that established the intelligence infrastructure for the healthcare economy, announced today the successful results of a 120-day clinical intervention study conducted in partnership with Dallas Endocrinology. The study validates the company’s core thesis: that “Agentic AI” can drive meaningful clinical outcomes and sustained engagement without the manual friction plaguing general-purpose LLMs like ChatGPT.

These intervention results underscore the efficacy driving MediKarma’s growing commercial footprint, validating the decision by partners like Family Health Choice to deploy 20,000 units of the platform to close care gaps at scale.

Beyond the Chatbot: Solving the “Blank Cursor” Problem

While the tech world focuses on the capabilities of ChatGPT, patients attempting to use it for health management face a massive hurdle: intensive labor. To get value from a generic LLM, a patient must essentially become a data engineer – manually locating, downloading and uploading scattered PDF records – and then possess enough medical literacy just to prompt the AI correctly.

MediKarma renders such manual friction obsolete.

• The MediKarma Difference: In under 120 seconds, MediKarma’s “frictionless ingestion” engine connects to any provider, retrieves a patient’s full clinical history, and instantly builds a personalized health profile.
• The Result: While ChatGPT waits for a prompt, MediKarma’s “Jill” agent is already proactively nudging the patient based on clinical truth established in the first two minutes of onboarding.

Intervention Study Results: The Dallas Endocrinology Cohort

The 120-day intervention deployed “Diabetes Jill”, an autonomous agent designed to act as a 24/7 extension of the care team.

• 50% Clinical Efficacy: Half of the participants successfully reduced their A1C levels or maintained a healthy baseline (A1C < 5.9).
• 35% Sustained Engagement: The platform achieved a 35% Monthly Active User (MAU) rate, significantly outperforming the single-digit engagement typical of standalone health apps.
• 100% Adoption: Every participant in the intervention cohort actively engaged with the agent.

Dr. Rohit Dwivedi, M.D., a specialist in Endocrinology, Diabetes & Metabolism and MediKarma Advisory Board member, commented on the metabolic significance:

“In metabolic care, the battle isn’t won in the clinic; it’s won in the thousands of micro-decisions patients make daily. Generic AI can recite guidelines, but it lacks clinical context. MediKarma’s ‘Agentic AI’ succeeds because it is contextually grounded in the patient’s actual medical history from day one. It doesn’t just answer questions; it anticipates the metabolic drift that physicians typically can’t see until it’s too late.”

Commercial Impact: Validating the Family Health Plan Partnership

The efficacy demonstrated in this intervention study reinforces the strategic value behind MediKarma’s major commercial deployments. Family Health Choice, a South Florida-based Prepaid Health Clinic, previously partnered with MediKarma to deploy 20,000 units of Wellness Jill, selecting the platform specifically for its ability to automate engagement without member friction.

Carlos Perez, CEO of Family Health Choice, highlighted the operational impact of the technology:

“We selected MediKarma for our 20,000-member deployment because they solved the engagement problem that other tools couldn’t. Most digital health solutions create more work for our members; MediKarma automates the ‘last mile’ by proactively guiding them between visits. Based on the platform’s ability to drive sustained interaction, we are targeting significant improvements in HEDIS gap closure and reductions in administrative costs, outcomes that would typically require a massive expansion of our human staff.”

Executive Commentary

Dr. Scott Howell, Chief Clinical Officer of MediKarma (Colonel, USAF Ret.), added:

“For AI to move from a novelty to a necessity, it must be frictionless. We didn’t just build a better chatbot; we built a clinical instrument that works immediately. While others expect patients to be data engineers for their own health records, we do the heavy lifting in 120 seconds. This intervention study proves that when you remove the friction, you unlock the outcomes.”

Looking Ahead: Next-Generation Behavioral Science

MediKarma views these initial engagement metrics as the floor, not the ceiling. The company is currently developing next-generation efficacy and engagement nudges based on advanced behavioral science frameworks. These enhancements are expected to further boost MAU rates and drive even deeper glucose improvements in future iterations of the platform.

About MediKarma

MediKarma is the operating system for the new healthcare economy. We provide the industry’s only patented, frictionless EMR ingestion engine (“Jack AI”) and a suite of validated, white-label AI agents (“Jill AI”) that empower health plans and systems to deliver personalized, preventive care at scale.

Media Contact:
Nataliya Howard
[email protected]
925.628.7424
www.medikarma.com

View original content:https://www.prnewswire.com/news-releases/medikarma-validates-agentic-ai-in-clinical-intervention-study-demonstrates-50-a1c-reduction-proves-friction-free-alternative-to-chatgpt-302689839.html

SOURCE MediKarma Inc.