Building in stealth — Spring 2026

The AI agent that keeps cardiac patients in recovery

800,000+ heart patients drop out of cardiac rehab every year. Resilio is the AI-powered engagement layer that works alongside your existing program — keeping patients accountable between sessions, guided by AHA protocols, at a fraction of the cost.

Request Early Access How it works

Completions are the whole game. For patients and hospitals.

Every patient who drops out is a missed reimbursement for the hospital and a worse outcome for the patient. The data is clear — completion is dose-dependent, and the gap between 1 session and 36 is the difference between life and death.

CMS opened the door in 2026: virtual cardiac rehab is now permanently on the Medicare telehealth list with virtual physician supervision. The reimbursement rails are built. The bottleneck is keeping patients engaged every single day.

Resilio is the AI engagement layer that works between your existing rehab sessions. It handles the daily touchpoints — check-ins, exercise nudges, symptom screening — that determine whether a patient finishes or disappears. Your clinical team keeps doing what they do best. Resilio makes sure patients show up.

$4,320

Medicare reimbursement per completed 36-session patient. At 74% dropout, hospitals forfeit ~$3,200 per eligible patient.

CMS 42 CFR §410.49; AACVPR 2023

47%

Lower mortality for patients completing all 36 sessions vs. those attending just one. Each additional session reduces cardiac readmission by 1.8%.

Hammill et al., Circulation (2010); Pack et al., JAHA (2013)

$1,005

Lower annual Medicare expenditure per CR participant — driven by 48 fewer hospitalizations per 1,000 patients per year.

Ritchey et al., JAHA (2024)

$1–5

Resilio's cost per patient to provide daily engagement across the full 36-session program. Augments your clinical team without adding headcount.

An AI coach grounded in clinical protocols

Resilio fills the gaps between rehab sessions — keeping patients engaged, accountable, and on track through all 36 sessions.

Clinical Feedback Loop

Every exercise session is evaluated against AHA/AACVPR protocol targets. Heart rate, duration, perceived exertion — the agent gives specific, grounded feedback referencing the patient's week-by-week progression.

Red-Flag Escalation

The agent never dismisses symptoms. Chest pain, shortness of breath, dizziness — any red flag triggers immediate clarification and clinical team notification. Safe by design.

Behavioral Nudges

Built on motivational interviewing and nudge science. When a patient skips a walk, the agent doesn't lecture — it asks what got in the way and offers a smaller alternative. Sustained behavior change, not compliance.

Built by the people who live this problem

A clinical cardiologist who watches patients disappear after discharge, and an AI product leader who builds the systems to catch them.

Sanjit Misra

CEO · Technical Co-Founder

Principal Technical PM at AWS, where he led accuracy and ranking systems for Amazon Q Business — managing 40+ engineers and applied scientists building agentic AI retrieval. Previously built clinical data products at Stanford Health Care, Navigating Cancer, and Zus Health across a 15-year healthcare technology career. MBA from Duke Fuqua. Dual BS in Chemistry and Mathematics from UNC Chapel Hill.

Agentic AI Healthcare Product AWS · Stanford · Duke

Satish Misra, MD

CMO · Clinical Co-Founder

Clinical cardiac electrophysiologist at Advocate Health — the largest nonprofit health system in the US, spanning 67 hospitals across 6 states. Hopkins-trained. Performs ablations and cardioversions weekly, writes cardiac rehab referrals, and watches patients disappear from follow-up.

Cardiac EP Advocate Health Johns Hopkins

Early preview