I architect multi-agent pipelines, LangGraph workflows, and intelligent backends — turning hours of manual work into minutes of automated precision.
AI Engineer with hands-on experience building production-grade agentic systems, RAG pipelines, and multi-agent orchestration using LLMs. I reduce manual workflows from hours to minutes through LLM-powered automation.
My focus is LangGraph orchestration, ReAct-based agents, integrating MCP servers, and human-in-the-loop system design — the gap between AI research and AI that reliably runs in production.
BE in Computer Engineering, Sagarmatha Engineering College, Nepal (2021–2025).
An AI interview platform where users upload their CV and a target Job Description. A RAG agent evaluates fit — eligible candidates proceed to an adaptive AI interview; others receive structured CV improvement suggestions.
Agentic RAG system for multi-turn document Q&A with persistent session memory. ReAct loop via LangGraph, semantic search with pgvector, dual chunking strategies, and dual embedding options.
Model Context Protocol server exposing web search as a callable tool for AI agents. Deployed on Render, compatible with Claude Desktop and MCP Inspector — plug-and-play across any MCP workflow.
Backend for a pet adoption platform. RESTful API built with Django Rest Framework, user authentication, and secured endpoints using Django's built-in auth system.
Open to remote opportunities, freelance AI projects, and interesting collaborations. Response within 24 hours.