Hire Gopi N. - Fine-Tuning Engineer

Worked on building and deploying an AI-powered knowledge assistant to support banking operations and compliance teams. I designed a Retrieval-Augmented Generation (RAG) system that combined secure document ingestion, semantic search, and vector-based retrieval to deliver accurate and policy-aligned responses using large language models. I developed a multi-agent workflow using LangGraph to manage routing and secure integrations with internal systems, ensuring reliable and compliant outputs in a regulated environment. In addition, I built and deployed a LightGBM-based predictive model to support operational risk analysis and case prioritization. I also implemented MLflow-based model tracking, monitoring, and cloud deployment processes to ensure scalability, performance, and long-term system reliability

I contributed to the development of an AI-powered Grant Writing Assistant designed to help researchers discover relevant funding opportunities and draft stronger proposals. I built a Retrieval-Augmented Generation (RAG) pipeline that combined semantic search with large language model responses to improve the accuracy and relevance of generated content. Using vector databases such as Qdrant and ChromaDB along with optimized embeddings from Sentence Transformers, I enhanced retrieval precision and alignment with funding agency requirements. I developed backend services using FastAPI and WebSocket to enable real-time interactions and ensured the system scaled to support 300+ concurrent users with high availability. On the frontend, I worked with React and Next.js to create an intuitive interface for proposal drafting and grant matching. Through retrieval tuning and re-ranking strategies, we significantly reduced manual writing effort and improved proposal relevance based on user feedback.

Worked on building scalable, AI-driven enterprise search and integration systems using Python (Flask), Vertex AI, and Google Cloud. I developed REST APIs, optimized natural language query handling with few-shot learning, and implemented intelligent autosuggestions to improve search accuracy and user experience. My work modernized legacy systems by migrating C# modules to Python, enhanced data retrieval speed and precision, and streamlined Azure DevOps workflows for efficient deployment and collaboration across teams.