LLM evaluation suite for Doctor AI
A RAG-based evaluation framework that tests Ultrahuman's Doctor AI feature and catches regressions, hallucinations and drops in answer quality.
~/projects
Projects
A selection of things I have built and worked on.
Self-healing web and mobile automation
Web and mobile UI automation that uses AI and MCP to find elements and repair itself, reducing the maintenance that test suites normally need.
Player automation for live streaming at scale
Automation for video playback at Disney+ Hotstar across web, Android and iOS players, on the squad responsible for streaming quality for 28M+ concurrent users during peak live events.
End-to-end QA for Blood Vision
Quality engineering for an AI feature that reads blood test reports, from PDF ingestion, through AI analysis, to what the user sees in the app.
API automation suite and AI test data
A Karate-based API automation suite that runs in Docker, with an AI-driven system that generates consistent, realistic user data for staging and QA.
Test observability dashboards
Company-wide Grafana and Prometheus dashboards showing real-time test analytics and Appium failure tracking for every squad.
Automation for a WebRTC conferencing SDK
Web, mobile and API automation for the 100ms live video conferencing SDK, including Flutter-based mobile automation frameworks.
College engineering projects
Three semester projects at R. V. College of Engineering, including leading a team through the end-to-end build of a hardware-based solution.