Waymark, the AI-first community clinic for Medicaid, today announced the launch of Waymark Compass, an SMS-based AI care navigator that helps patients enrolled in Medicaid get assistance with their care, benefits, and community resources. Compass is supervised by Waymark physicians and built on peer-reviewed algorithms that have been rigorously tested by Waymark’s community-based teams.
More than 70 million Americans rely on Medicaid, where limited access to timely primary care and social services leave patients vulnerable to an avoidable emergency room visit. A wave of consumer AI health tools now offers medical guidance directly to patients, but peer-reviewed research has repeatedly found that general-purpose models miss real emergencies and route patients with urgent conditions to routine follow-up. Many of these tools also answer questions but do not provide actual direct services to patients.
Available in more than 25 languages (covering over 99% of the languages spoken by patients in Medicaid), Compass lets patients get instant help with their care, benefits, and community resources, with Waymark physicians supervising every clinical interaction. The tool is designed to take action on behalf of patients, such as scheduling appointments, arranging transportation, and connecting with their care team directly. Compass is delivered over SMS with no app or login required, and is built specifically to ensure clinical safety for Medicaid patients: trained and clinically tested on real patient cases, physician-supervised, and engineered to detect emergencies using multiple clinical triage frameworks.
Waymark's prior research showed its community-based care model reduced acute care use by more than 20 percent compared to a control group (NEJM Catalyst, 2024). In a peer-reviewed study of more than 3,000 Medicaid patients, Waymark's AI-guided care recommendations reduced acute care events by 20.7 percent compared with community-based teams only following care management protocols— with no cases of worsened outcomes attributable to following the AI's recommendations (JMIR AI, 2025). Compass extends that same evidence-based approach directly to patients by pairing a language model with a separate, rule-based clinical layer that checks every interaction against established medical guidelines It is designed to catch the two primary issues that undermine general-purpose models: errors of omission, where a model gives a confident answer but misses an urgent diagnosis or needed test, and errors of fabrication, where a model invents a condition, treatment, or guideline that doesn’t exist.
"When a Medicaid patient texts at 2 a.m. asking whether a symptom is serious, a confident guess isn't good enough and a wrong answer can mean a missed emergency," said Sanjay Basu, MD, PhD, Co-Founder and Chief Medical and Technical Officer at Waymark. "We built our tools so a language model handles the conversation but a rule-based, physician-supervised layer makes every clinical decision, and that decision can be traced and explained."
Compass is the latest addition to Waymark’s proprietary AI platform, which also includes the Waymark Signal Suite of predictive models. Unlike traditional risk models that identify patients after crisis patterns emerge, Waymark Signal proactively predicts who will benefit from intervention before an emergency room visit or hospitalization, and recommends what interventions to provide.
Waymark pairs this technology with multidisciplinary, community-based care teams — including community health workers, care coordinators, pharmacists, and therapists — who reach rising-risk patients enrolled in Medicaid before their health needs become emergencies. Compass extends that model directly into patients’ hands as a new tool for 24/7 support, giving every patient a concierge for their care between visits. Waymark will continue to make Compass available to eligible patients across its markets in the coming months.
