Tanay Gaur

GenAI Engineer with 4+ years shipping production LLM and multi-agent systems — trade finance document automation, agentic RAG pipelines, and voice AI applications. Deep hands-on experience with LangGraph, LangChain, FastAPI, and vector databases (Pinecone, Qdrant, ChromaDB); built and deployed systems processing 500+ documents/day and cutting manual workflows by 40–60%. Currently pursuing an MSc at the University of Winnipeg; available part-time or contract (remote) for GenAI engineering roles

Product Engineer at Intellect Design Arena (2025-09 – 2026-07)

• Built a full-stack AI-powered trade finance workbench (React + FastAPI) processing 100+ LC documents per day via parallel Purple Fabric agent execution — covering OCR, classification, structured extraction, TBML compliance, and discrepancy detection; async UI and OpenShift deployment reduced manual review time by ~40%
• Appointed as Product Owner and architected a multi-agent, LLM-based system for automated test case generation from unstructured specification documents, implementing advanced prompt engineering, recursive document parsing, and orchestration logic that improved model inference accuracy and structured output quality across diverse, unstructured document types

AI Engineer at Launch Ventures (2025-02 – 2025-08)

• Built an automated newsletter creation tool using LangGraph state management and GPT-4, integrating real-time web research (Tavily API) and image curation (Google CSE), cutting editorial production time by ~50%
• Developed a context-aware FAQ chatbot (Pinecone + LangChain + OpenAI + Crawl4AI) that auto-crawled and indexed website content, reducing support response latency by ~60% and enabling automated, context-grounded answers for end users
• Built a Document Analyzer allowing users to upload and query large files (PDF, CSV, DOC), enabling efficient extraction and analysis across diverse document formats for 10+ enterprise clients
• Developed an old-age companion for active seniors enabling chat and voice interactions, medical concern search, and cognitive games, leveraging Grok (Llama 3.3 70b), Tavily, PostgreSQL, ChromaDB, Livekit agents, Deepgram, and ElevenLabs

Software Engineer at Persistent Systems (2022-06 – 2024-05)

• Engineered a RAG++ pipeline for business and litigation data using hybrid Vector + Knowledge Graph indexes, improving information retrieval precision by ~35% and enhancing entity and relation extraction with LLMs
• Awarded cross-functional team recognition for outstanding performance on the RAG++ pipeline implementation — cited for technical execution, collaboration, and measurable retrieval improvements on business and litigation data
• Developed the Agentic Copilot — an intelligent agent system (RAG + LangChain + Qdrant + LangFuse) that streamlined task assistance, optimized information retrieval, and provided comprehensive logging for cutting average task resolution time for 3 cross-functional engineering teams of 10+ members

Machine Learning Intern at Persistent Systems (2021-12 – 2022-06)

• Developed data preprocessing and model evaluation pipelines in Python that fed directly into production RAG and LLM retrieval systems — early foundation for agentic AI work at Persistent Systems

MSc in Applied Computer Science and Society – University of Winnipeg (2026-09 – 2028-08)

BTech in Computer Science – Lakshmi Narain College of Technology (2018-07 – 2022-07)