Shivang Rai

I help organizations design and ship data platforms that turn complex source systems into usable analytics. With a background spanning B.E. Mechanical and M.Sc. Chemistry, I transitioned into data and AI to build solutions that are both technically solid and business-focused.

At Thorogood, I worked on building a data platform from scratch for West Star Aviation, including extracting transformation logic from Power BI dataflow JSON, implementing a Medallion Architecture (Bronze/Silver/Gold), and creating a metadata-driven framework to ingest and transform 900+ tables from 5+ sources using shared Databricks notebooks. I also orchestrated data pipelines with Databricks Jobs to integrate into Power BI semantic models, and improved Power BI report performance by optimizing the semantic model, removing unused rows, moving transformations into Databricks, and creating pre-calculated tables.

I further supported large-scale Unilever analytics by processing accounting-related datasets with PySpark, automating monthly reporting via Azure Data Factory into Excel extracts, and redesigning ADF pipelines to reduce overall failure risk. Alongside consulting, I built Python-based revenue automation and ad analysis systems, and explored machine learning for CNC milling output prediction using multiple supervised models.