Ojaswini Shrivastav

I’m a Data & Analytics Engineer with 2+ years of experience building production ETL pipelines, enterprise analytics dashboards, and AI-powered automation solutions. I focus on translating business problems into actionable insights through data.

At Chubb Business Services India, I designed and maintained production incremental ETL pipelines on Databricks (PySpark) using watermark-based incremental loading, recursive JSON flattening, and window functions to calculate 20+ business KPIs. I deliver curated datasets to Snowflake via Delta Lake, supporting reliable reporting and downstream decision-making.

I build analytics products that teams can use day-to-day, including production Power BI and Qlik Sense dashboards to monitor STP, claims performance, automation rates, renewal efficiency, and operational KPIs. I also developed Prodigy, an AI-powered SQL generation assistant using GPT-4o, LangChain, and Chainlit, and implemented an enterprise Retrieval-Augmented Generation (RAG) solution using OpenAI GPT, FAISS, LangChain, and LangGraph.

Before that, I completed internships building predictive and computer-vision models, including a YOLOv5 label detection model and regression models in Python. I enjoy working cross-functionally to define requirements, validate KPIs, and ship solutions that hold up in production.