One dashboard.
Six departments.
Built to scale.
An eCommerce company needed a single source of truth across their entire operation. I built the data infrastructure to make that possible, and structured it so it works for however they grow.
Details have been redacted to protect client confidentiality.
The Problem
Six departments. Ten-plus data sources. No shared visibility.
Every team was pulling their own numbers from Shopify, Amazon, ad platforms, and customer tools, then maintaining their own Google Sheets on top of it. That's 1 to 2 hours per department, per week, just to keep the reports alive. Multiply that across six teams and you have a part-time job dedicated entirely to knowing how the business is performing. And even then, nobody had the full picture.
Each platform had its own reporting interface. No standardized format. No single place to ask how the business is doing today. Every department was working in isolation, with no way to connect the dots across sales, marketing, CX, supply chain, finance, and ops at the same time.
The Solution
One pipeline. Zero manual input. Built once.
The goal was one pipeline that handles everything without anyone on the team having to touch it — and built so it doesn't need to be rebuilt as the business grows. Every tool in this stack was chosen on purpose.
Tech Stack & Data Flow
Once it's set up, nobody on the team has to touch it. The data moves, the reports update, and every department starts the week with the same numbers.
The Medallion Architecture
This is where most BI setups go wrong.
Most BI builds put the data model inside the BI tool itself — which means the tool and the data become inseparable. Switch tools later and you start from scratch. The data was structured inside Databricks using a three-layer approach specifically to avoid that. The BI tool is just a window into the data. The data itself lives independently.
Databricks is also not just a dashboard engine here. The same data foundation is already set up for what comes next — data enrichment, deeper analytics, and eventually an AI layer that can answer business questions directly from the company's own data. That was part of the plan from day one.
The Delivery Layer
One page. Every report. Any browser.
I built a custom HTML page that consolidates all the Power BI reports into one place. Each report is embedded directly into the page, so every department sees their numbers in a single view without jumping between tabs or tools. The page is hosted on Netlify with password protection, using the company's own domain to keep the branding consistent. Any team member can access it from a browser with no licenses and no friction.
What Changed
The spreadsheets are gone. The manual pulls are gone.
Six departments now work from the same numbers, updated automatically every week. And the time each team spent on reporting every week compounds into real savings the longer the pipeline runs. That's the point of building infrastructure right the first time.
The infrastructure is also already set up for what comes next — whether that's data enrichment, deeper analytics, or an AI agent built on top of the same foundation. That was never an afterthought. It was part of the decision from day one.