Syntameter AI-Powered Speech & Language Assessment

Role: Lead Fullstack & Mobile Engineer Tech: React, React Native, FastAPI, Python (spaCy), ASR/OCR pipelines, PostgreSQL, AWS Overview

Syntameter is an AI-driven platform that automates speech transcription, syntactic scoring, and language-progress analytics for teachers, therapists, and researchers. I led the full build of both the web dashboard and the mobile app, enabling fast, on-device audio capture and real-time scoring.

Teachers and therapists were manually transcribing speech samples and calculating syntactic scores — a process taking hours per student and impossible to scale across classrooms.

I designed and developed a cross-platform system consisting of a React dashboard and a fully native-feeling React Native mobile app. Mobile users can record speech samples, upload audio, view instant scoring, and track progress. Core engineering work included:

React Native Mobile App – Audio recording, offline-first caching, progress visualization, and fast uploads.

AI Scoring Engine – spaCy dependency parsing, utterance segmentation, sentence-type classification, and syntactic complexity scoring.

IRB-Compliant Backend – FastAPI microservices, secure media processing, and consent workflows.

Live Analytics – D3.js / Chart.js visualizations with cross-device syncing.

Reduced assessment time by 85% (from hours to under 90 seconds).

ASR segmentation accuracy improved to >90% across supported languages.

Mobile adoption reached 70% within 3 weeks due to ease of use.

Enabled multi-school rollout with secure data workflows and parent-facing reports.