Ashwani kumar

I’m a Senior Data Engineer with strong expertise in designing and implementing scalable data architectures. I focus on leveraging big data technologies to optimize data workflows and improve data accessibility, so teams can make better data-driven decisions.

At GEP, I created the SDM(v2) data pipeline in PySpark and Scala-Spark to handle end-to-end ingestion, cleansing, consolidation, profile reporting, transliteration (via ChatGPT), and quality checks. The Spark code processed over 300 million transactions (~500 GB), and I built an ADB automation test pipeline for regression testing. I also owned the release process through a CI/CD pipeline, optimized performance using persist, broadcast joins, accumulators, and partitioning, and implemented medallion architecture and timely maintenance (optimize, vacuum) to reduce infra cost.

Previously at OPTUM - UHG, I worked as an Analyst on healthcare cost prediction using machine learning and statistics—covering models like regression, gradient boost, LightGBM, and random forests. I automated pipelines and delivered outputs to US stakeholders, generating revenue worth $500K, and I led a $300K “Group Cost Predictors” project from India while driving analytical feature engineering and insights.