• 🎓 Third-year Computer Science student at St. Joseph's College of Engineering, Chennai
• 🔭 Currently building AI-powered applications — an AI-assisted SOC/forensics platform, face-recognition tracking systems, and full-stack learning platforms
• 🌱 Deepening my skills in LLMs, Retrieval-Augmented Generation, and Computer Vision
• 🏆 Building ASPIRE — a planned multi-role CCTV student tracking system (InsightFace-based), started at a hackathon
• 💡 Open to AI/ML and Full-Stack Developer internship opportunities
• 📫 Reach me on LinkedIn or Email

Languages

Frontend & Backend

AI / ML / LLM

Tools & Infra

A multi-tenant SOC workbench that combines email header/authentication analysis, AI-driven threat scoring, graph-based campaign correlation, and tamper-evident evidence generation for incident responders.

• 🔐 Multi-tenant workspace isolation with server-side RBAC (Owner/Admin/Analyst), JWT auth with refresh-token rotation, and fail-closed cross-tenant access
• 🧪 Full analysis pipeline: RFC 5322 header parsing, SPF/DKIM/DMARC validation, IOC extraction, SHA-256 attachment hashing, and LLM-based threat scoring with a strict AI/forensic-evidence separation
• 🕸️ Neo4j campaign correlation graph linking related emails, domains, IPs, and hashes, enriched via VirusTotal, AbuseIPDB, and IP geolocation
• 📄 Tamper-evident PDF evidence reports with cryptographic (SHA-256) custody chains
• 🔌 Custom MCP server exposing 16 investigation tools (header parsing, IOC extraction, enrichment, graph queries) to AI agents like Claude Desktop and Cursor
• ⚡ Real-time updates via Redis pub/sub + SSE, background processing on Celery workers
• 🔜 Google Workspace/Gmail sync and Shodan enrichment are partially wired up; Merkle-tree proofs and a decentralized evidence ledger are on the roadmap

📎 Repository · 🔗 Live Demo

Hackathon project (Makeathon) · Phase 1 covers student registration and face-embedding capture; multi-camera detection, tracking, and role dashboards are planned for Phase 2.

• ✅ Built: student registration, webcam face capture, 512-d ArcFace embedding generation and storage (SQLite)
• 🔜 Planned: multi-face CCTV tracking (YOLOv11 + ByteTrack), role-based dashboards (Admin/Mentor/Staff/Student), QR-based logging, department-level analytics
• 🎯 Designed as the scaled-up successor to my earlier SmartAttend project below

📎 Repository

A full-stack, multi-user learning platform built around my own 200-Day Challenge: a structured curriculum from DSA through Core/Advanced Java, DBMS, System Design, and production-grade engineering.

• 🗺️ 200-day curriculum spanning DSA (Days 1–70), Java Core & Advanced/Spring (71–110), DBMS (111–130), System Design (131–160), and capstone projects (161–200)
• 🔐 Multi-user isolation with JWT auth; per-account tasks, streaks, notes, and metrics
• 🏆 Live leaderboard ranked by XP, completed tasks, and streak consistency
• 📊 Weekly mastery charts and track-level progress analytics
• 💬 Markdown-enabled daily notes and a peer discussion forum
• 🧪 Automated test suite covering auth, multi-tenant isolation, and curriculum integrity

📎 Repository · 🔗 Live Demo

Admin-facing attendance system with single-face-per-frame detection and a live analytics dashboard.

• 👤 Single-face detection per frame with automatic attendance marking
• ✍️ Manual correction option for admins to fix misreads
• 📊 Analytics dashboard: attendance trends, department comparisons, low-attendance alerts

📎 Repository · 🔗 Live Demo

An Instagram-style platform for campus — post events, pitch ideas, raise funding, and collaborate on projects with other students.

• 📢 Post and discover college events with registration built in
• 💡 Share project ideas and raise funding/support for them
• 🤝 Collaborate with other students on shared college projects
• 🗓️ Central feed for everything happening on campus

📎 Repository · 🔗 Live Demo

A personal AI chatbot built on a Retrieval-Augmented Generation (RAG) pipeline for context-aware conversations.

• 🧠 Retrieves relevant context before generating responses (RAG architecture)
• 💬 Personal-use chat assistant with grounded, context-aware answers

📎 Repository

⭐️ Thanks for visiting my profile — feel free to explore my repositories!