Developed a computer-aided diagnostic system for brain tumor (MRI) and pneumonia (X-ray) detection. Implemented CNN and Vision Transformer models in TensorFlow with preprocessing for denoising and segmentation. Built Flask APIs with JWT authentication and deployed Dockerized services on Render for scalability. Achieved 92% and 89% accuracy respectively, enabling real-time inference and confidence scoring for clinicians.

Constructed a scalable web app for URL shortening with custom encoding and click analytics. Enhanced RESTful API endpoints, incorporated SOAP and third-party APIs, and supported 1,000+ users.

Launched an AI-driven meeting summary platform automating note generation and task tracking. Delivered sub-2s response latency via Vercel edge deployment and NeonDB optimization. Facilitated 500+ active users across multi-tenant workspaces.

Engineered an AI-powered expense management platform automating receipt scanning, categorization, and reporting. Realized 90% data accuracy and reduced manual expense processing by 50%. Supported 100+ active users with serverless APIs on Next.js and PostgreSQL.

Built a file-sharing platform using the MERN stack with AWS S3 integration for secure file storage, enabling users to upload, download, and share files via unique links with customizable access controls

Built a real-time chat app with Next.js 14, Next Auth, MongoDB, Pusher, React-Hook-Form, Tailwind CSS, and Next Cloudinary for secure messaging, authentication, and image handling.