Sreeja Suragouni

I'm building LLM-powered banking assistants at DE Capital, automating customer inquiries, email triage, and knowledge-base search through production RAG pipelines. I use LangChain, Pinecone, domain-specific embeddings, and secure FastAPI inference services to improve retrieval quality and operational reliability.

Previously at Liberty Mutual Insurance, I built fraud detection, claim severity, and forecasting models using Python, SQL, XGBoost, LightGBM, and scikit-learn. I deployed ML workloads on AWS SageMaker and built CI/CD workflows that standardized packaging, testing, and deployment.

Across Tata CLiQ and Merck Group, I've delivered customer segmentation, demand forecasting, recommendation experiments, clinical analytics, ETL workflows, and business dashboards. I translate complex data problems into practical ML solutions using Python, Spark, SQL, AWS, and production-focused MLOps practices.