Production BI copilot running on local LLMs (Ollama) so business users ask questions in plain English instead of waiting on an analyst.
40+ business users · sharp cut in analyst dependency · near-zero recurring cost
Director of Analytics, BI and AI — built data functions from zero across D2C, e-commerce, marketplace and financial services. Based in Bengaluru, open to roles in BLR, Pune, UAE & Europe.
Analytics, BI and AI leader with 12+ years building data functions from zero across D2C, e-commerce, marketplace and financial services, including 6+ years leading the function end to end.
Built the Analytics and BI function at Uppercase from pre-revenue to ₹140 Cr annually, redesigning scattered system extracts into a single, query-optimised warehouse across 20+ sources. Owned analytics across web, mobile app, marketing, product and operations data at Flipkart and 6thStreet, running personalisation, product-ranking and seller-churn models in production. Hands-on with SQL, Python, Snowflake, Power BI and Tableau, with a track record of partnering directly with Category, Finance, Product and Executive leadership.
Production BI copilot running on local LLMs (Ollama) so business users ask questions in plain English instead of waiting on an analyst.
40+ business users · sharp cut in analyst dependency · near-zero recurring cost
Anomaly detection across margins, pricing, inventory and working capital, flagging outliers before they show up in a weekly review.
80% faster issue detection
Single source of truth built from SAP, OMS, POS, DMS and marketplace extracts into query-optimised flat tables, replacing ticket-driven reporting.
20+ systems unified · ~40 hrs/week saved
SKU-level P&L tracking across pricing and inventory, used to rationalise the catalogue and cut wasted spend.
₹20L+ in quarterly savings
Gender-prediction model built from browsing and purchase behaviour to close the gap in female customer representation.
grew traffic and revenue share from a targeted persona
Behavioural scoring model to flag marketplace sellers likely to go dormant, prioritised for re-engagement.
~400 sellers reactivated on a base of 6,000