Wearable-AI™ metabolic intelligence that runs where life happens.

Detect metabolic risk earlier. Intervene before damage begins.

More than 1.3 billion adults worldwide are metabolically unhealthy — and roughly half are unaware of their risk. Metabolic disease progresses silently, from glucose dysregulation to insulin resistance, organ damage and chronic disease, and today's healthcare typically detects it only after irreversible damage has begun. NeuTigers builds compact deep-learning models that read metabolic signals from the devices people already own — smartwatches, wearable sensors and connected health data — and deliver metabolic intelligence in real time, directly on the device: early risk signals, where and when they can still change the outcome.

How it works

1. Multi-sensor wearable data. Everyday wearables capture rich physiological signal — heart rhythm and pulse, blood pressure, body composition, oxygen saturation, skin temperature, motion, activity and sleep — combined with lifestyle questionnaires and clinical data under rigorous data-quality and privacy controls.
2. On-device AI engine. Our StarDeep platform converts that data into compact, edge-optimized neural networks using Princeton-origin methods for model efficiency, data quality and synthetic-data augmentation — transforming raw sensor signals into digital biomarkers of metabolic health, computed on the device itself.
3. From risk prediction to prevention. The resulting models are packaged into disease-specific programs — such as SweetDeep™ for type 2 diabetes risk and MFDeep for women's metabolic health — that turn digital biomarkers into risk predictions and personalized, actionable insights driving early action and prevention, each developed with clinical partners and carrying a clearly stated evidence status.

Explore the platform · See our health programs

What partners gain

• Differentiated metabolic-intelligence features, faster — build on a working model-development pipeline and existing metabolic programs instead of starting from raw research.
• Reduced cloud dependence — inference runs on the device, cutting latency, connectivity requirements and operating cost.
• Privacy by design — physiological data can be processed locally rather than streamed to remote servers.
• Continuous real-world data — everyday wearables become a source of longitudinal metabolic signal, within appropriate consent and governance frameworks.

Explore partnership opportunities · Review clinical and scientific evidence

Program spotlight

• SweetDeep — our flagship program, turning every wearable into a diabetes risk scanner: non-invasive, real-time type 2 diabetes risk screening with a smartwatch. Prospective study with AP-HP; arXiv publication, 2026.
• MFDeep — the women's-health pillar of our metabolic portfolio: fetal macrosomia risk linked to gestational diabetes, delivered on-device to medical teams. In active clinical development with Genopole and CHSF, supported by France 2030.
• StarDeep™ Agentic AI Platform — a multi-agent platform for secure, intelligent and scalable decentralized clinical trials, powering the studies behind our programs.

Each program carries an explicit maturity label; none implies regulatory clearance. See the full portfolio.

Our vision

We are building the metabolic intelligence layer for every wearable. Today, that means SweetDeep and diabetes risk; tomorrow, a broader cardiometabolic platform; in the near future, AI health intelligence systems for preventive care. The objective stays constant throughout: by putting clinically grounded metabolic intelligence on devices people already own, NeuTigers works toward making earlier detection and prevention accessible and equitable, globally — wherever you live, whatever clinic you can reach.