Innovative and results-driven Generative AI Architect with extensive experience in LLM customization, fine-tuning, and prompt engineering to address complex business use cases. I have a strong track record of building and deploying cutting-edge conversational AI and RAG-based chatbots across multiple domains, enhancing user engagement and delivering high-value solutions. With deep expertise in BERT, NER, and LLMs such as Llama2, Mistral, and Falcon, I consistently tailor models to achieve exceptional precision and performance. Key Achievements: Enterprise RAG Platform: As a ML Consultant, developed a conversational AI RAG platform using LangChain and LlamaIndex, implementing a ReAct Agent to automate workflows and improve query handling efficiency. Financial Data Chatbot: Built a RAG-based chatbot for financial data at Apollo Learning, using document chunking, embeddings, and CosmosDB to deliver accurate query results. Applied metadata filtering and security compliance, improving the chatbot's reliability and precision. Legal Billing Automation: Fine-tuned the Llama2 model for legal billing validation, achieving a 25% reduction in billing rejections through enhanced model accuracy, deployed with Docker and SageMaker for scalable AI solutions. Foundational LLM Research: Led R&D efforts on fine-tuning Llama2, Mistral, and Falcon models, tailoring them for specialized use cases like legal billing prediction and creditworthiness assessment from business news data. My background as a C/C++ developer has been instrumental in delivering high-performance solutions for leading companies, including: BMC Software iSeries Server Monitoring: Developed APIs to optimize CPU utilization tracking in virtual environments. Telecom Content Filtering: Engineered an Apache Traffic Server plugin that handled over 1,000 concurrent users for secure and efficient content filtering. I’m passionate about advancing LLM-based solutions that deliver impactful results, elevate user experiences, and drive the future of AI-driven innovation.