The AI engine behind our programs
Replica Health is an AI/ML research lab, charting the path to safe and effective AI for biosensor data analysis and automated drug delivery. Everything below was built for our own programs — and is available to partners.
Software we've built
MetaboNet
The "ImageNet for diabetes" — the largest standardized, consolidated Type 1 diabetes dataset: 2,030 patient-years across 3,135 subjects and 21 source datasets, with tiered public and DUA-governed access at metabo-net.org. It is also the foundation for MetaboNet-Bench, a multi-modal glucose-forecasting benchmark developed in collaboration with Stanford.
CarbCast
From a meal description and context, predicts the full carbohydrate-absorption curve in the format insulin-delivery systems accept. Evaluated head-to-head against human carb estimates in the Tidepool × Stanford study presented at ADA 2026.
EventSense
Recognizes meals and activities from on-device context — CGM trend, time of day, location, and activity — enabling one-tap logging. Runs fully on-device with over-the-air model updates.
RapidUVAPadova
A re-engineered, massively parallel UVA/Padova simulator — 14× faster in-silico testing across virtual patients, with a Python interface for new control policies, including multi-GPU workloads.
MetaboNet-Replay
Recreates real patient trajectories from MetaboNet, then replays them under new insulin-delivery policies — a grounded counterfactual that de-risks algorithms before they ever touch a patient.
Replica-Capture
An iOS app plus serverless backend that powers real-world data collection for clinical trials — capturing diabetes, activity, and location data straight into the MetaboNet format.
Our Capabilities
Research
Replica is an independent R&D lab available to medical device companies, medical data companies, and pharmaceutical companies. We offer unparalleled data collection capabilities for use in clinical trials via our iOS App.
Integration
Whether you are a medical device or pharmaceutical company, Replica provides API access and bespoke software components to meet your company's metabolic analysis needs.
iOS App
Replica's AI-powered diabetes management app reveals personal carb absorption patterns and more. Replica iOS lets patients understand the relationship between their lifestyle and blood glucose control.
Partner with us to transform care
Our Partners & Affiliations
Our Team
Sam brings extensive expertise in machine learning and predictive analytics. In academia, Sam's research focused on incorporating location and activity into blood glucose predictive models. Before Replica, Sam co-founded and led engineering at an anomaly detection startup.
Miriam holds a PhD in Computer Science from NTNU, specializing in evaluating blood glucose prediction algorithms, and is an active open-source contributor.
Nat is an expert at machine learning on constrained devices — ex-Google TensorFlow team and founding engineer at Moonshine AI. A lifelong Type-1 diabetic and Stanford visiting scholar in metabolic simulation and control.
Courtney, a former Software Engineer at Virta Health, brings expertise in developing software interventions for Type-2 Diabetes and now focuses on Replica's mobile and front-end development.