At MTBC CareCloud, I develop AI-powered healthcare solutions, agentic AI systems, and scalable microservices. My work includes a medical document indexing system that extracts and auto-populates patient and billing data, achieving 90%+ extraction accuracy.
I also developed a multi-agent medical coding platform using hierarchical RAG, multi-corpus retrieval, fine-tuning, and search-based pipelines. Its microservices architecture includes a rules engine, and the platform achieved approximately 80% coding accuracy in production-oriented healthcare workflows.
For VibeSync, my final-year project, I built an AI system that recommends background music by analyzing video mood and content. I developed deep learning pipelines for feature extraction and emotion recognition, along with a MERN web platform for user interaction and music suggestions.
My other projects include an LSTM air pollution prediction model using live sensor data, a CNN-based CAPTCHA OCR system, and a dual-input CNN for COVID detection using X-ray and CT data. I was runner-up at the FAST-NUCES AI Ideafest for VibeSync.
