Demand-Sense Forecasting Engine
SKU-level demand forecasts across 30,000 items, blending LightGBM with regional weather and promotion signals. Cut stockouts 22%, reduced excess inventory by $3.4M.
Senior Strategy Intelligence Scientist & Data Engineer ACI Limited
I turn messy enterprise data into the signal that drives boardroom decisions — building data platforms, forecasting models and decision-grade dashboards across pharma, agri, consumer, retail and logistics businesses.
From the raw warehouse layer to the executive dashboard — these are the tools, methods and frameworks I lean on day to day.
A snapshot of platforms, models and decision systems I've designed and shipped — most under NDA, summarised here at a high level.
SKU-level demand forecasts across 30,000 items, blending LightGBM with regional weather and promotion signals. Cut stockouts 22%, reduced excess inventory by $3.4M.
Unified identity graph stitching 4.2M customer records across pharma and consumer business units.
One-screen command centre tracking margin, working capital and SLA across 14 markets — refresh every 15 min.
Replatformed legacy ETL stack to a Databricks lakehouse — cut nightly batch from 6h to 38m.
Causal-uplift model for the consumer business — A/B validated, lifted gross margin by 3.1 points.
Streaming pipeline for 1,200 vehicles — anomaly detection surfaces in under 30 seconds.
From frontline analytics to platform leadership — eight years of compounding work.
Lead the Strategy Intelligence pod serving six business units. Built the executive forecasting & scenario-planning practice from scratch and own the lakehouse roadmap.
Shipped the company's first production ML platform on Azure + Databricks; productionised demand and pricing models across the consumer and pharma businesses.
Owned reporting and KPI design across retail and after-sales service. Replaced 40+ spreadsheet reports with a self-serve Power BI workspace adopted by 200+ users.
Joined out of university as the team's first dedicated analyst. Built the warehouse foundation, taught myself SQL on the job, and never looked back.
I'm always happy to talk shop — strategy data, ML in the enterprise, or your next interesting problem. Drop a line, I read every message.