Case Study · product
AutiSense – Autism Screening Platform
Comprehensive pre-autism screening product for kids aged 2–6 pairing deep CNN computer vision with clinician dashboards.
Track
product
Technologies
Dart, Flutter, React JS, Flask, TensorFlow, PostgreSQL
Primary Impact
Pre-autism screening for ages 2–6
RepositoryView on GitHub ↗
Context & Challenge
Early detection of childhood developmental conditions is often hindered by long clinical waitlists and inaccessible computational tools.
Process & Implementation
Trained a convolutional neural network on 5,000+ behavioral and facial image samples achieving 86%+ classification accuracy; served predictions via Flask REST API with sub-200ms latency; built mobile and web clinician interfaces backed by PostgreSQL.
Verified Outcomes & Key Facts
- Pre-autism screening for ages 2–6
- CNN on 5,000+ labeled samples
- 86%+ classification accuracy
- sub-200ms REST inference latency