Projects

Open-source terminal AI coding-agent harness for studying and building reliable agent loops, schema-first tools, MCP integrations, skills, subagents, persistence, checkpoints, safety boundaries, and a Rich terminal UI.

MCP-native organizational intelligence server that turns Crustdata-style company, hiring, people, and web evidence into cohort briefs, company phase classification, and structured go-to-market insight for Claude.

A production-ready RAG-powered document intelligence platform that enables users to upload documents, chat with AI using intelligent retrieval, and generate content with proper citations. Features hybrid search, reranking, and secure document processing.

Desktop AI assistant built with Electron, React, TypeScript, Mastra, MCP, and Composio to reduce context switching with local chat, screen capture, voice input, and workflow automation.

Artificial Guruji is an AI-powered exam preparation platform that uses advanced AI to generate personalised study materials and customised study schedules. It helps users to improve test scores, adapt to individual learning styles, and save preparation time. The platform has over 50,000 users and is used in over 100 universities.

GuardianReport is a system for secure and anonymous incident reporting that uses military-grade encryption to protect the reporter's identity. It provides a two-way anonymous communication channel with law enforcement, and offers real-time processing of reports. The system has a 100% anonymity rate, has processed over 100,000 reports, and offers 24/7 support.

Built a responsive and visually appealing landing page for a music institution using Next.js and Acerenity UI. This project demonstrates expertise in creating clean, modern designs with seamless navigation and optimized performance.

Built a Vision Transformer (ViT) model to efficiently classify images using self-attention mechanisms. Implemented with PyTorch and fine-tuned on a curated dataset, achieving high accuracy. This project demonstrates proficiency in transformer-based architectures and advanced deep learning techniques for visual tasks.