A universal data model for a market that has no standard

Deregulated energy markets do not agree on anything. Suppliers, distributors and brokers operate under different regulations by region, and every supplier emails its price matrix in its own format. The job was to make all of that queryable.

I worked with the business analysts to scope it, which meant reading actual regulations covering energy suppliers, distributors and brokers across unregulated regions, and then designing a data model general enough to hold all of it.

Built so the boring part could be outsourced

The ETL system takes supplier-specific adapters, each one converting a daily email of matrix files into our database. That was a deliberate architectural choice: isolating the per-supplier quirks into small adapters meant writing new ones did not require touching the core, and the work could be handed off.

The database model tracks the origin of every price down to the source file and line number. QA analysts could then answer “where did this number come from” without guesswork, which in a regulated market is not a nice-to-have.

The application on top

I built the Django web application where an energy broker searches and filters prices, then walks a workflow to produce quotes and visualizations for their own customers and prospects. That included a reporting dashboard, a small template engine, and PDF export driven by headless browser printing from the command line.