Sindhu Raghuram Panyam

Master's

Data Science

Pinnacle AI

AI ML Engineer

- Engineered a multimodal GraphRAG using - Neo4j, LlamaIndex, and Langchain, and prompt engineered - AWS Bedrock - Anthropic Claude v3.5, enhancing the creation of complex service documentation from audio, visual and textual data - Leveraged Advanced RAG with Metadata Filtering, Hybrid Search, LLM re-ranking, and in-context learning, to improve the response relevancy by 26% and drive customer satisfaction - Incorporated streaming integration with the UI using FastAPI and NextJS, reducing the response time from 10s to under 2s - Processed and clustered conversational data stored in Amazon S3, extracted semantically rich FAQ's for customer support - Designed and implemented a playlist expansion Recommendation System, experimenting with and optimizing 3 Graph - Neural Network architectures, deploying the optimal model and contributing to 17% of the projected annual revenue - Generated and uploaded the Docker image to Amazon ECR and used Fargate container orchestration - Deployed the applications to AWS Lambda