At Mahavir Timber, I analyzed transactional data across 23 SKUs to support inventory and pricing decisions. I automated the cleaning and consolidation of sales, inventory, and dispatch records, then developed a K-Means pipeline to group SKUs by regional profitability.
I delivered interactive Power BI dashboards that turned the findings into recommendations for purchasing and stocking decisions. My project work includes AttributionIQ, a graph-based marketing attribution pipeline, and ContractSense, a contract-risk chatbot using Hybrid RAG.
I also built MarketPulse AI to analyze earnings narratives and generate risk-adjusted position recommendations. For my YouTube channel analytics project, I examined more than 1,000 channels and found a 100x geographic CPM gap and a statistically significant content-dilution effect.
