At Groove Innovations, I contributed to developing and deploying a production RAG chatbot for a US-based construction client. It helped users query 200+ architectural PDFs with source-grounded answers and page-level citations.
I'm developing backend services for Quantixone, a multi-tenant CRM platform, using Python and FastAPI. My work includes AI agents, client management, subscriptions, forms, and media storage.
For Quantixone, I designed and implemented an event-driven notification service using AWS EventBridge, SNS, SQS, Lambda, SES, and S3. It supports asynchronous delivery, status tracking, retries, and dead-letter queue processing.
As a freelancer, I developed a real-time human action recognition pipeline using YOLO-Pose keypoint extraction and LSTM-based sequence modeling. I also developed a Video RAG project for multimodal question answering and content summarization across video.
