Full Stack Developer
XCEED, NIT JalandharILEED
Intelligent Learning Engagement and Entity Detection
140 PRs · 131 merged · Jun – Sep 2026
- Real-time student identification from live RTSP streams and video uploads, on a FAISS vector index with RetinaFace (ONNX) detection.
- Automated ERP face-embedding generation, nightly embedding rebuilds, and active learning from students' backup photos.
- Moved RTSP recording to the Node.js backend: scheduled recordings, FFmpeg audio extraction, and live-preview auto-retry.
- One-click deploys with health checks and automatic rollback, a GPU dashboard with MIG/VRAM telemetry, and uptime alerting.
2.0s → 915ms
ERP list load
11.3%
timeout error rate resolved
91% heap
stream memory leak fixed
XCEED Learning App
Android & iOS app for the AMS platform
291 commits · 5 PRs · Aug – Sep 2026
- Built the app from the first commit on Capacitor: PIN and biometric sign-in, deep links, and multi-account switching.
- Delta over-the-air updates published through GitHub Actions, shipping only the files a release changed.
- Automated web → app code sync with custom merge drivers, keeping mobile-only code behind override seams.
- Push notifications with Firebase Cloud Messaging and offline-first caching with TanStack Query.
21.6 → 0.41 MB
per small OTA release
75
automated web → app syncs
Android + iOS
from one codebase
Organisation repositories are private; numbers come from my authored PRs and commits.