Early-Stage Development • TRL 4

Building trust in non-invasive glucose measurement.

A reliability-first approach to optical glucose monitoring, focused on safety, uncertainty handling, and clinical validation. Designed for the long road to regulatory approval.

SWIR
Optical Approach
Glucose-relevant spectral regions
Safety
First Priority
Qualified output over false confidence
TRL 4
Development Stage
Lab validation in progress
See Development Demo Meet the Founder
124
mg/dL
Stable
FDA 510(k) Ready
SWIR Spectroscopy
Real-time AI
The Problem

537 million people live with diabetes. All of them hate needles.

🩸

The Pain Problem

Current CGMs require a needle insertion every 10-14 days. Over a diabetic's lifetime, that's 3,000+ sensor insertions. The constant trauma causes scar tissue, reduced accuracy, and psychological burden.

⚠️

The Accuracy Gap

Today's best invasive CGMs achieve 7.8% MARD. FDA allows up to 15%. Every inaccurate reading is a potential insulin dosage error. Lives depend on better accuracy.

💰

The Cost Burden

A year of CGM supplies costs $4,000-6,000. Insurance gaps leave millions without access. A permanent, reusable solution changes the economics entirely.

🏥

The Clinical Limitation

Hospital glucose monitoring still relies on fingersticks. Limited integration with clinical systems. No unified platform exists for enterprise healthcare glucose management.

"Every technological leap in medicine has been about reducing invasiveness. We did it for surgery. We did it for imaging. Glucose monitoring is next."
Our Solution

We read glucose through light, not blood.

💡 SWIR Light Emission
↓
🔬 Tissue Penetration
↓
📊 Spectral Analysis
↓
🧠 AI Inference
↓
✅ Clinical Output

The Nephora S1 System

01

Multi-Wavelength SWIR Spectroscopy

Multiple short-wave infrared wavelengths operate in glucose-relevant spectral regions. Early validation shows promise for non-invasive detection.

02

Qualify-Then-Measure Paradigm

A core design principle: qualify tissue conditions before measurement. The goal is to reduce noise and improve reliability. Development results show improved signal quality.

03

Measurement Intelligence Kernel

Our software kernel is designed to prioritize safety and reliability. When conditions are uncertain, the system withholds output rather than provide false confidence.

04

Enterprise-Ready Platform

Architecture designed with healthcare integration in mind. FHIR R4 compatibility planned for future clinical deployments.

Development Demo

System behavior illustration. Simulated data for demonstration.

Signal Processing Pipeline

📡
SWIR Sensors
Multi-Wavelength SWIR
→
🔧
Signal Processing
Lock-in Amplifier
→
🧠
ML Inference
Feature Extraction
→
🛡️
Safety Gate
Uncertainty Check
→
📱
Output
mg/dL + Confidence

Nephora S1 - Simulated Reading (Not for clinical use)
118
mg/dL Confidence: 94%
● In Range
24-Hour Trend
6 AM 12 PM 6 PM Now
Session Metrics
87% Time in Range
112 Avg Glucose
5.8% GMI (est. A1C)
18% Variability
Design Philosophy

A different approach to glucose measurement.

Conceptual comparison to illustrate design intent. Not a performance-equivalent claim.

Aspect Invasive CGMs (Dexcom / Abbott) Nephora S1 (Development System)
Measurement Method Subcutaneous electrochemical sensing Optical SWIR spectroscopy (research-stage)
Tissue Interaction Penetrating sensor External, non-invasive
Sensor Replacement Periodic (10–14 days) No consumable sensor (hardware reuse)
Initialization Warm-up required post-insertion Optical stabilization period (under study)
Accuracy Metrics Clinically validated (FDA-cleared) Validation pending (targets defined)
Failure Handling Continuous output Output gating under uncertain conditions
Development Stage Commercial Pre-clinical (TRL 4)
The Hardware

Purpose-built for non-invasive glucose sensing.

Nephora S1 Finger-Clip CGM Device Nephora S1 Exploded View

Hover to explore internals

Nephora S1

Clinical-grade accuracy without the pain

📡
SWIR Optical System Multi-wavelength SWIR LEDs
🔬
High-Sensitivity Photodetector High-sensitivity spectral sensing
🧠
Embedded Inference Processor On-device ML inference
📊
High-Precision ADC Medical-grade precision
The Technology

Built for the FDA. Designed for doctors.

🌐

Client Layer

Doctor Portal Mobile App Device Dashboard API Gateway
🏥

Enterprise Layer

FHIR R4 API EHR Sync OAuth2/JWT Audit Logging
⚖️

Regulatory Layer

Calibration Engine Safety Monitor FDA Compliance ISO 15197
🧠

Machine Learning Layer

Model Registry Feature Extraction Uncertainty Engine Drift Detection
📡

Signal Processing Layer

Digital Lock-In Gain Control Noise Rejection Coherent Detection
⚡

Hardware Layer

SWIR LEDs Precision Photodetector High-Performance MCU Low Energy Wireless

Development Targets

  • Target MARD: < 10% (validation pending)
  • Clarke Zone A+B: Target 95%+
  • Measurement Range: 40-400 mg/dL
  • Update Rate: Every 5 minutes

Signal Processing

  • ADC Resolution: High-Precision
  • Sampling Rate: High-Frequency
  • SNR: Clinical Grade
  • Lock-In Time: Sub-second

Regulatory Design

  • FDA 21 CFR 820: Designed for
  • IEC 62304 Class B: In progress
  • ISO 14971: Planned
  • ISO 13485 QMS: Planned
Defensibility

Five moats. Each compounds over time.

01

Data Moat

Every device logs every measurement with consent. Ground truth collection from fingerstick comparisons. Per-user, per-population, per-condition segmentation.

Competitor Response Time: 3-5 years
02

Calibration Moat

Hardware fingerprinting + continuous calibration eliminates drift. Every device has a unique signature. The fleet learns together – issues in one device inform all devices.

Competitor Response Time: 2-3 years
03

Verification & Auditability Moat

Full traceability from input to output. Uncertainty logging at every stage. Reproducible measurement chains. Designed for regulatory scrutiny and clinical validation.

Competitor Response Time: 4-5 years
04

Evolution Moat

Safe over-the-air model updates. Canary deployments. Shadow mode testing. Automatic rollback. While competitors ship static models, our accuracy improves continuously.

Competitor Response Time: 3-4 years
05

Regulatory Moat

Built-in compliance infrastructure. IEC 62304 baked into development. ISO 14971 integrated. 21 CFR Part 11 from provenance chain. First-mover regulatory relationships.

510(k) Cost Savings: 60%

The Compounding Flywheel

🔄 Compound
Growth
📊 Data Moat
🧠 Better Models
🛡️ Trust Moat
✅ FDA Approval
👥 More Users
📈 More Data
Intellectual Property

Novel approaches designed for defensible IP protection.

4
Key Innovations Identified
3
Core Technology Areas
Full
Stack Coverage

Optical Sensing Approach

Novel methodology for non-invasive physiological measurement using spectral analysis

Intelligent Signal Processing

Adaptive measurement paradigm with qualification-first architecture

Reliability-First Output Control

Confidence-aware decision system designed for clinical safety

Leadership

Meet the Founder

Ujjwal Manot - Founder & Director, Nephora Health

Ujjwal Manot

Founder & Director

I founded Nephora Health with a singular conviction: the era of painful glucose monitoring should have ended long ago. The fundamental technologies exist SWIR spectroscopy, advanced signal processing, machine learning yet no one has integrated them correctly.

We are not developing yet another CGM device. We are building the definitive platform that enables truly non-invasive continuous glucose monitoring. Every component, from the optical hardware to the regulatory compliance infrastructure, has been engineered from first principles.

This is not a research initiative destined to remain in the laboratory. Nephora Health is a company designed to deliver real products to clinicians and patients. Our patents are being prepared for filing. The software architecture is complete. The entire system is designed for FDA approval.

"I am here to end the needle era. This is not merely a tagline it is our mission."

🎯 Developed complete technology stack from hardware to cloud
📄 4 patent applications to be filed covering core IP
🏥 Designed for FDA 510(k) / De Novo pathway
What We've Built

This isn't a pitch deck. This is working technology.

💻

Full Software Stack

  • Production-ready FastAPI backend
  • React/TypeScript doctor dashboard
  • Real-time WebSocket telemetry
  • Device monitoring system
🧠

AI/ML Pipeline

  • Feature extraction engine
  • Model registry with versioning
  • Uncertainty quantification
  • Drift detection system
🔐

Enterprise Features

  • FHIR R4 API implementation
  • JWT/OAuth2 authentication
  • Role-based access control
  • Complete audit trail
⚡

Hardware Foundation

  • SWIR optical system design
  • Signal processing algorithms
  • Digital lock-in amplifier
  • Hardware abstraction layer
210+ Source Files
87 Documentation Pages
31 Test Modules
4 Core Innovations
Market Opportunity

$20.7 billion by 2030. We're capturing the fastest-growing segment.

$8.3B
Global CGM Market (2024)
$20.7B
Projected Market (2030)
15.8%
CAGR

Addressable Segments

Consumer Wearables Largest Growth
Hospital Systems Enterprise Revenue
Insulin Pump Integration Clinical Necessity
The Vision

We're not building a product. We're building healthcare infrastructure.

Phase 1: Clinical Validation

100-patient clinical trial. Refine algorithms. Establish regulatory pathway.

Phase 2: Regulatory Submission

FDA De Novo pathway. Complete verification and validation testing. Pursue clearance.

Phase 3: US Commercial Launch

Partner with health systems. Enterprise sales. Consumer launch.

Phase 4: Global Expansion

CE marking (EU). CDSCO (India). NMPA (China). Global platform.

Building the foundation for trustworthy non-invasive glucose measurement.

537 million people need better options. We're taking the long road: rigorous validation, safety-first design, and honest uncertainty quantification.

Early-stage. Reliability-focused. Designed for the regulatory journey ahead.

"Output only when conditions justify it. Validation before scale."