GEN AI Developer

•Created AI-powered solutions using AI agents and LangGraph to convert natural language into SQL queries, enhancing data accessibility and reducing query processing time by 50%. •Led the development of an LLM-powered Retrieval-Augmented Generation (RAG) system using LangChain as the orchestrator and Azure OpenAI services, enabling users to query complex census documents (HTML, PDFs with tables, and text) leveraging Unstructured IO to parse the content, resulting in a 40% increase in query accuracy. Built and deployed AI solutions leveraging Azure AI services (e.g., Cognitive Services, Azure ML, Azure AI Search). •Deployed ML models using AWS Bedrock and SageMaker; automated infrastructure with Terraform and managed version control via Git. Leveraged Kubernetes for efficient container orchestration, cutting deployment time by 25%. •Developed and deployed neural network models, including LSTM-based models, to enhance predictive capabilities and performance of AI solutions, achieving a 20% improvement in prediction accuracy. •Fine-tuned LLMs using Supervised Fine-Tuning (SFT) and Low-Rank Adaptation (LoRA) techniques, achieving a 60% increase in model accuracy. Additionally, fine-tuned and implemented NeMo Guardrails to prevent out-of-topic conversations, jailbreaking attempts, and abusive language, reducing such incidents by 70%. •Applied Model Context Protocol (MCP) to streamline development processes and improve code quality, reducing code review time by 15%.

Bachelor of Technology

Microsoft Certified Azure AI Developer

Azure AI104