I'm a B.Tech Computer Science student focused on Data Science who likes turning ideas into working software. My projects explore agentic AI, LLM-powered applications, workflow automation, computer vision, and robotics.

My main interest is in systems that go beyond a text response: retrieving context, calling tools, keeping state, and connecting models to real workflows.

• 🤖 Agentic AI: tool use, multi-agent workflows, RAG, streaming, and memory
• ⚙️ Automation: APIs, MCP integrations, and event-driven utilities
• 👁️ Computer vision: detection, tracking, and image/video processing
• 🦾 Robotics: simulation and industrial automation concepts

AI systems · agents · retrieval

An enterprise-intelligence prototype built around RAG and decision logic. It brings together streaming chat, Gmail analysis, persistent memory, and source-aware context to help turn fragmented information into suggested next actions.

A modular chatbot with local LLM support, PDF RAG, tool calling, remote MCP integrations, persistent conversations, and streamed responses.

A multi-agent pipeline that researches a user-provided topic, synthesizes findings, writes a conversational script, and generates narrated audio.

Computer vision · applied AI

Real-time object detection and identity tracking for webcam or video input using YOLOv8 and Deep SORT.

An LSTM-based music-generation pipeline that learns patterns from MIDI datasets and composes new sequences.

A browser-based 3D voxel editor with real-time hand-gesture controls, built with vanilla JavaScript and Three.js.

Automation · real-world utilities

An AI-assisted civic issue reporting concept with photo/GPS reports, duplicate detection, road-health maps, and resolution workflows.

Compares timetable snapshots and sends Telegram alerts only when meaningful class details change, retaining the old baseline when delivery fails.

A read-only Moodle watcher that sends Telegram alerts for assignment deadlines, announcements, course resources, and notifications.

A voice-controlled desktop music assistant combining speech recognition, speech synthesis, and Spotify playback.

Languages

AI / ML

Backend / tooling

• More reliable agents with clear tool boundaries, persistent state, and observable workflows
• Grounded RAG systems with traceable sources
• Voice-first assistants and real-time human–AI interaction
• Robotics, simulation, and perception systems

_{Stats cards are provided by an external service and may take a moment to load.}

Open to collaborating on thoughtful AI, automation, and developer-tool projects.