Punithsai Chokki

About Punithsai Chokki

I’m a data engineer with 5+ years of experience building data platforms that actually move the needle — from powering fraud detection at scale to making healthcare systems more efficient. I’ve worked across AWS, Azure, and GCP environments, and I enjoy designing pipelines that are clean, fast, and reliable — whether they’re processing 10 million transactions or helping a team make decisions in real time. Right now at CITI Bank, I’m focused on building real-time streaming pipelines and fraud analytics solutions using Azure Databricks, Synapse, and Cosmos DB. Before that, I helped Mercury Insurance cut $2M in fraudulent claim payouts by building a cloud-native analytics platform that combined ML scoring, SageMaker, and AWS Glue to deliver insights in seconds. At Fortis Healthcare and Glenmark Pharmaceuticals, I led data engineering efforts around patient journey analytics and global supply chain intelligence — experiences that taught me a lot about building secure, high-impact data solutions in regulated environments. I’m comfortable across SQL, Python, Spark, and PySpark, and I’ve worked on both real-time and batch pipelines. I’ve built out dimensional models, developed in CI/CD environments using Azure DevOps and Jenkins, and collaborated with cross-functional teams to turn business needs into technical deliverables that scale. Whether it’s streaming sensor data or integrating complex reporting across cloud services, I enjoy the problem-solving side of data engineering — and the satisfaction of seeing those solutions in action. I’m always looking to learn something new and enjoy working with teams that value quality, collaboration, and thoughtful engineering. If you\'re working on something interesting or just want to connect over shared challenges in the data world, feel free to reach out. I’m always up for a good conversation.

Experience

1. Azure Data Engineer

   Citi

   Apr 2024 — Present · Wilmington, IL, US

   Engineered and launched a real-time fraud detection system using Azure Databricks, Synapse, and ML models, cutting fraud response time by 40% across more than 10 million monthly transactions. Developed robust ingestion and transformation pipelines with Azure Data Factory and PySpark, processing high-volume batch and streaming data from SQL, Cosmos DB, and Kafka with sub-5-second latency. Deployed ML models through Azure ML and TensorFlow to identify anomalous behavior, improving precision by 20%; streamlined deployments using Azure DevOps, Jenkins, and Kubernetes to reduce release time by half.

Education

• University of the Pacific - Eberhardt School of Business

  Master of Science - MS