Computer Science undergraduate specializing in Computer Vision, MCP-based AI systems, and agentic tooling and applied AI systems. Building cognitive vision software and real-time intelligent pipelines for edge hardware..
Currently, I’m exploring advanced AI tooling, MCP (Model Context Protocol), and Claude-based agentic systems, with a strong interest in building intelligent systems that combine vision, reasoning, and reliable execution under pressure.
My work focuses on deploying reliable, real-world Computer Vision and Edge AI systems rather than experimental prototypes. During my time with the Neurology Department at All India Institute of Medical Sciences
I learned that model accuracy alone is meaningless if a system cannot sustain stable inference on constrained hardware in clinical environments. I design and optimize robust pipelines using OpenCV, custom neural networks, and Raspberry Pi 4/5 edge deployments with an emphasis on low latency, hardware efficiency, and real-time performance.
B.Tech in Computer Science & Engineering
A production clinical website built for the Neurology Department at AIIMS New Delhi during my internship. Covers neuroscience research, clinical excellence, and patient care workflows. Built to support real hospital operations — not a prototype.
Built with OpenCV and Haar Cascade classifiers to detect eye regions and estimate gaze direction from live webcam feeds. Optimized for Raspberry Pi 4/5 for edge deployment and applied in cognitive and behavioral monitoring at AIIMS.
Automated identification system using Python and OpenCV to record attendance with timestamps from live video. Includes robust face encoding, matching logic, and data logging engineered for real-world classroom environments.
A game controlled entirely by hand gestures using MediaPipe Hands for hand landmark-based gesture recognition, validation, and live score tracking — a fully interactive Human-Computer Interaction application.
A cloud-hosted MCP-powered GitHub analytics system that analyzes developer profiles, repositories, programming languages, and coding activity using the GitHub REST API. Built with Python and FastMCP, the platform delivers AI-driven insights, repository intelligence, learning trend analysis, README reviews, and developer profiling ...
An AI research assistant leveraging Model Context Protocol and Claude to query, synthesise, and summarise academic content — built for clinical and technical research workflows.
A registry for managing and discovering MCP-compatible tools and extensions. A Tool which is used by the developer and normal user too.enabling developers to publish, discover, and install Model Context Protocol servers via a live searchable platform with real-time like counts, category filters, and one-click install command copy, backed by a REST API and persistent database.
Led cybersecurity workshops and technical sessions for students. Mentored junior members on network security fundamentals and threat analysis, building a culture of hands-on security awareness.
Winner of multiple college-level debate competitions. Active in public speaking and technical discussions, bridging the gap between complex AI concepts and general audience understanding.
Whether you want to discuss edge AI deployment, have a research opportunity, or just want to talk computer vision — my inbox is always open.
amanprataps218@gmail.com