Nidhi Madnani

AI Engineer with 2+ years of experience building scalable Generative AI applications, retrieval systems, and agentic AI workflows for enterprise financial platforms. Specialized in production-ready RAG architectures, multi-agent systems, AI copilots, and intelligent automation using LangChain, LangGraph, Azure AI Foundry, and modern LLM ecosystems. Proven track record taking AI products from POC to production in high-impact enterprise environments.

Software Engineer at Indus Valley Partners, Pvt. Ltd. (2026-01 – Present)

• Architected enterprise-scale AI document intelligence and semantic retrieval solutions for hedge fund front-, middle-, and back-office workflows, enabling natural-language access across thousands of enterprise documents.
• Designed intelligent metadata enrichment and document tagging pipelines using enterprise reference data to power Azure AI Search, improving retrieval precision for operational queries.
• Contributed to building an enterprise knowledge graph and ontology layer on Microsoft Fabric, establishing contextual relationships across documents, workflows, and operational datasets.
• Built and deployed Copilot-based conversational AI agents via Microsoft Copilot Studio, Azure AI Foundry, and Azure AI Search, rolled out across Microsoft Teams and SharePoint channels.
• Developed MCP (Model Context Protocol) integrations connecting internal wikis, case management systems, and enterprise knowledge repositories to multi-agent orchestration frameworks.
• Engineered specialised retrieval agents with persistent memory and semantic search over 200K+ historical enterprise cases, improving contextual reasoning and historical issue resolution speed.

Associate Software Engineer at Indus Valley Partners, Pvt. Ltd. (2024-01 – 2025-12)

• Designed and optimised end-to-end RAG pipelines for credit agreement processing using LangChain and LangGraph — including contextual retrieval, custom chunking, and prompt-engineered entity extraction — improving retrieval and response accuracy from ~70% to ~90%.
• Built AI-assisted financial statement processing workflows using Hugging Face models and Azure Document Intelligence for structured financial extraction, chart-of-accounts mapping, and quarter-over-quarter data quality validation, achieving near-perfect extraction accuracy.
• Developed and optimised backend APIs and enterprise services using Django and Django REST Framework (DRF) to support AI-powered document processing and workflow orchestration systems.
• Mentored 3 junior developers on AI engineering concepts including RAG pipelines, vector databases, chunking strategies, prompt engineering, and agentic workflows.

Master of Computer Applications in Computer Applications – NMIMS Mukesh Patel School of Technology and Engineering (MPSTME) (2022 – 2024)

Bachelor of Science in Information Technology – Jai Hind College (2019 – 2022)