Looking for a job

Sree Vaishnavi Rao

@sreevaishnavirao

AI Software Engineer at Magellan Health building HIPAA-compliant clinical RAG and multi-agent workflows for behavioral health.

United States

Message

At Magellan Health, I architect clinical RAG pipelines over healthcare documents and engineer multi-agent workflows for prior authorization and clinical note summarization. I also fine-tune language models and deploy HIPAA-compliant inference pipelines.

At McKinsey & Company, I developed Generative AI advisory interfaces and RAG workflows for enterprise knowledge retrieval. Earlier, at Dell Technologies, I built device-management APIs and microservices and developed machine-learning models for hardware failure prediction.

Experience

Work history, roles, and key accomplishments

Current

AI Engineer

Current

Magellan Health

May 2026 - Present (6 months)

Architected clinical RAG pipelines using AWS Bedrock, LangChain, and Pinecone over 100K+ healthcare documents, enabling HIPAA-compliant knowledge retrieval for behavioral health decision support. Engineered multi-agent LLM workflows using LangGraph and Claude 3 Sonnet APIs to automate prior authorization and clinical note summarization, reducing manual case processing time.

AWS Lang Chain Pinecone LangGraph Claude 3 Sonnet QLoRA PEFT Hugging Face Transformers AWS Sagemaker Amazon EKS MLFlow Prometheus Ragas Prompt Engineering

AI Software Developer

McKinsey & Company

Jul 2025 - Apr 2026 (9 months)

Developed React.js and TypeScript interfaces for internal Generative AI advisory platforms, integrating real-time LLM response streaming via Server-Sent Events. Spearheaded enterprise RAG pipeline development using LangChain and Pinecone, indexing 6M+ consulting documents and cutting knowledge retrieval latency.

React TypeScript Lang Chain Pinecone D3.Js LangGraph OpenAI API Python fastAPI PyTorch Hugging Face Transformers

Software Developer

Dell Technologies

Jun 2021 - Aug 2023 (2 years 2 months)

Applied Python and Scikit-learn to analyze 320K device logs, building machine learning failure prediction models that identified recurring hardware patterns. Engineered service-to-service integrations across 5 internal platforms using REST APIs and RabbitMQ, reducing data synchronization latency.