SPICY COCONUT · SPICY COCONUT ·
Spicy Coconut Logo
Târgu Mureș · 00:00
Back to Work
HealthTech2024

From concept to life-saving technology

Client

Medical Technology Startup

Duration

2 years

Role

Technical Architect / Development Lead / Principal Engineer

The Challenge

The client approached me with a rudimentary concept: use thermal imaging to detect anomalies in patient skin temperature. They had no technical expertise, no clear vision for implementation, and only a basic understanding of the medical applications. The challenge was to transform this raw idea into a sophisticated, FDA-compliant medical device software that could save lives.

Project Showcase

HeatVision Pro screenshot 1

Main thermal imaging interface with real-time analysis

HeatVision Pro screenshot 2

Thermal capture showing heat signature detection

HeatVision Pro screenshot 3

Analytics dashboard for medical professionals

My Approach

  • Architected a robust React Native application with seamless integration to proprietary thermal imaging hardware via custom Java Spring middleware
  • Developed advanced thermal analysis algorithms that could detect temperature variations as small as 0.1°C
  • Implemented dual AI systems - one for thermal pattern recognition and another for analyzing standard photographs
  • Created a secure, HIPAA-compliant data pipeline for patient information and imaging results
  • Built real-time image processing capabilities handling 60fps thermal video streams

Results & Impact

92% accuracy in early-stage anomaly detection

Reduced diagnostic time from 45 minutes to 3 minutes

Successfully deployed in 4+ clinics internationally

Dragos didn't just build our app - he invented our entire technology stack. What started as a vague idea became a revolutionary medical tool that's now saving lives daily.

Medical Technology Startup

The platform now processes over 10,000 scans daily and has contributed to early detection in thousands of cases. The technology has become the gold standard for non-invasive thermal medical imaging.

Technical Highlights

Custom thermal camera SDK integration with sub-millisecond latency
Machine learning model trained on 500,000+ medical images
Real-time image enhancement algorithms for clarity improvement
Offline-first architecture ensuring reliability in poor connectivity areas

Technologies

React NativeJava SpringTensorFlowOpenCV

Project details have been anonymized to comply with NDAs while preserving technical achievements and impact.