Parth Bansal

Tech: React (TypeScript), Node.js, Express, MongoDB, Docker, Socket.IO

• Architected and developed a full-stack competitive coding platform supporting individual practice and team-based contests with real-time leaderboards and submission evaluation.
• Engineered a secure Docker-based sandboxed code execution engine with CPU and memory isolation to safely run multi-language programs (C++, Java, Python, C) at scale.
• Implemented JWT-based authentication, asynchronous submission queueing, and WebSocket-powered real- time updates to enable scalable and responsive contest execution.

Tech: FastAPI, LangGraph, PostgreSQL, Qdrant, Redis, React (TypeScript), Tailwind, Docker

• Built an AI-powered business analytics platform that analyzes e-commerce data and generates strategic insights using LLM-based research agents and retrieval-augmented generation (RAG).
• Designed a multi-step agentic workflow using LangGraph implementing intent classification, contextual memory retrieval, multi-pass analysis, and automated report generation.
• Implemented vector-based semantic search using Qdrant and sentence-transformer embeddings to enable contextual data retrieval from product catalogs, reviews, and competitor datasets.
• Developed a scalable full-stack system with an asynchronous FastAPI backend, PostgreSQL for structured business metrics, Redis caching, and a React dashboard with interactive analytics visualizations.

Contributed to UI/UX design for technical events, social media campaigns, and event branding; applied design principles to create engaging and user-centric visuals.