I'm a Backend & AI Engineer and CS junior at NIT Kurukshetra ('27), building at the intersection of scalable APIs, agentic AI systems, and Graph RAG. I care deeply about making AI actually useful — not just impressive demos, but production-grade tools that solve hard engineering problems.
- Currently building: LegacyGraph-MCP — an MCP server that exposes C++ codebase dependency graphs to AI agents using
tree-sitter+ Graph RAG. 20k+ RPC calls in two days. - Former: Backend Intern @ Lepton Software — engineered multi-agent AI assistant for conversational data analysis.
- Open to: Backend, Cloud, or AI Engineering internships for 2026.
- Currently exploring: LangGraph orchestration, vector search, distributed systems.
- CodeChef: 3-Star Rated Competitive Programmer.
Languages
AI & Backend Frameworks
Cloud & DevOps
AI Concepts
Repository Memory & Session Context Skill for AI Agents
A skill that scans repository history and conversation context to capture design decisions, open questions, and session state as plain Markdown — giving AI agents persistent project memory.
JavaScript AI Tooling Markdown Context Management
MCP Server for AI-powered C++ Legacy Codebase Modernization
Built a Model Context Protocol server that exposes C++ codebases as Knowledge Graphs to AI agents. Uses tree-sitter for 100% accurate AST parsing and NetworkX for circular dependency detection — solving the core problem of hallucinations in LLM-based legacy refactoring.
Python tree-sitter NetworkX FastMCP Graph RAG JSON-RPC GCP Smithery
End-to-End Document Intelligence Platform
An intelligent document analysis platform integrating OCR, RAG pipelines, and vector search, exposed through high-performance RESTful APIs deployed on Google Cloud Platform.
FastAPI RAG OCR Elasticsearch GCP Vector Search
Autonomous Multi-Agent System for Conversational Data Analysis
Built during internship — an agentic AI assistant powered by LangGraph that orchestrates multi-step reasoning across distributed APIs and databases. Enabled non-technical users to query complex data through natural language.
LangGraph LangChain FastAPI Multi-Agent PostgreSQL
I'm always open to discussing backend architecture, AI agent design, or just geeking out over distributed systems. If you're building something interesting or have an opportunity in mind — reach out!
