Open-Source Automated Insulin Delivery

Your insulin pump sees glucose and carbs.
Dossi sees the whole picture.

Dossi learns how sleep, exercise, meals, stress, and 50+ factors affect your insulin needs—then adjusts delivery automatically.

Join the Waitlist See how it works

50+

Factors Learned

5-Layer

Safety Chain

24/7

Closed-Loop Control

The Problem

Current pumps are reactive, not predictive

Blood sugar is shaped by dozens of factors—meal composition, sleep quality, stress, exercise, hormones, illness, even time of day. But current insulin pumps only see two inputs: carbs and glucose. They don't know you slept badly, had a high-fat snack, or just went for a run.

People with Type 1 spend only 30–40% of their day in a healthy glucose range. The problem isn't user error—it's that current systems are too simple.

Sleep

Poor sleep increases insulin resistance the next day

Exercise

Sensitivity increases for up to 48 hours after a workout

Meal Composition

Fat and protein cause delayed glucose spikes hours later

Stress

Cortisol drives hepatic glucose output, raising levels

Hormones

Menstrual cycle and dawn phenomenon shift needs daily

Illness

Inflammation-driven resistance can double insulin needs

How It Works

Intelligence that learns your body

Continuous Monitoring

Reads glucose every 5 minutes via Dexcom Bluetooth. Syncs sleep, exercise, and heart rate from Apple HealthKit—no manual logging.

Contextual Learning

A Bayesian engine learns your individual patterns across sleep quality, infusion site age, dawn phenomenon, exercise, and caffeine.

Predictive Intelligence

Models glucose trajectories 4 hours ahead. Quantifies uncertainty. Adjusts insulin delivery in real time through closed-loop control.

AI Nutrition

Snap a photo of your meal and Dossi identifies the food, estimates macros, and predicts glucose impact—replacing manual carb counting.

Dashboard Demo

Nutrition

Continuous Learning

Refines models with each new data point

After every meal, Dossi tracks what actually happened and refines its understanding of your unique physiology.

1Predict — Modeled glucose outcome

2Observe — What actually happened

3Attribute — Which factors were active

4Update — Refine personal model

By the Numbers

What's under the hood

Thousand Lines of Swift

50+

Factors Tracked

5-Layer

Safety Chain

Developer

Bayesian Learning

Learns individual glucose patterns across 5+ contextual factors. Only applies learned effects after 15+ observations with 50%+ confidence. Baseline drift detection triggers recalibration.

BLE Pump Driver

Native Bluetooth LE driver for Omnipod DASH. Pod pairing, encrypted sessions, status monitoring, basal adjustments, and bolus delivery—all built from scratch.

5-Layer Safety

Hard-coded limits, physiological bounds, hypo prediction, anomaly detection, and immutable audit logging. Biometric auth required for every dose. TOCTOU protection at delivery time.

AI Nutrition

Photo meal recognition via Gemini API. Snap a photo, identify foods, estimate macros, and predict glucose impact—no manual carb counting required.

Closed-Loop Control

Model predictive control optimizes dosing decisions continuously. Reads glucose every 5 minutes and adjusts basal rates in real time through a 90-minute optimization horizon.

Full Stack

SwiftUI + SwiftData locally, Supabase for auth and cloud sync, WidgetKit for home screen, watchOS companion app. Swift 6 strict concurrency throughout.

Capability	Traditional Pumps	Dossi
Current glucose
Carb counting
Exercise awareness
Sleep quality impact
Meal type (fat, protein)
Time-of-day patterns
Photo meal recognition
Stress & HRV awareness
Menstrual cycle tracking
Site age tracking

Your insulin pump sees glucose and carbs.
Dossi sees the whole picture.

Current pumps are reactive, not predictive