Chaithra Rajesh

Spearheaded the development of a customer-support chatbot, ensuring high accuracy and improved user satisfaction for internal and external stakeholders. Developed and deployed the first phase of a Document Translation Service using Azure Services with a unique capability to translate both textual documents and embedded images. Enhanced Retrieval-Augmented Generation (RAG) capabilities by integrating multiple document types, Confluence, and websites to improve chatbot accuracy and overall user experience. Orchestrated end-to-end migration to a serverless AI architecture on AWS Bedrock, reducing infrastructure costs by 40% and reducing model response time by 3-6 seconds. Led the in-house Live Chat CRM system, enabling seamless integration and real-time customer interactions.

Contributed to modernizing the HSEQ application by migrating legacy WPF interfaces to React and Tailwind CSS. Spearheaded development of Generative AI for a Chemical Assistant using llama 7b 4-bit quantized model and prompt engineering to validate viability. Built core components of a full-scale AI application, integrating Mistral 7b, Milvus vector database, and BAAI/bge-reranker-v2-m3 models while addressing GDPR and GPU scaling challenges. Developed and implemented a Dynamic Icon Management project frontend using Next.js.

Completed initial comprehensive training in software development fundamentals and best practices. Gained hands-on experience contributing to projects utilizing the React framework.