Case study · Application build · 2021–2022

Spikeet

A financial market data application built end to end, on top of an existing database covering 25 years of US stock data.

  • 25 yrsof US stock data
  • GBsscanned per screen
  • No-codefor non-technical traders

The engagement

Spikeet came to us with a substantial asset already in place: a database of 25 years of historical US stock data. What they needed was the application on top of it — the layer that would let traders actually use that data. We built that application end to end during a 2021–2022 engagement.

What we built

The core engineering problem was market-wide scanning: running complex, multi-condition filters and calculations across gigabytes of historical data, and returning results a trader could act on.

  • Market-wide scanning across the full 25-year dataset, not a sampled subset.
  • Multi-condition filters with calculations evaluated across gigabytes per scan.
  • A no-code interface, so traders could build and run screens without writing a line of SQL.

That last point defined the product. The audience was traders, not engineers — the power of the dataset only mattered if someone without a query language could reach it.

no-code screen 3 conditions 25 yrs · GBs matches
a screen anyone can build, a scan no spreadsheet could run

Scope, precisely

We built the application; the underlying database pre-existed our engagement. Our involvement ran through 2021–2022 and has since concluded, so this page describes the system as we delivered it — not the product as it stands today.

25 yrsof US stock data behind the application
GBsscanned per multi-condition screen
No-codescreen building for non-technical traders
  • ClientSpikeet
  • EngagementApplication build, end to end
  • Period2021–2022 (concluded)
  • DomainFinancial market data
  • DatabasePre-existing, 25 years of US stock data

Sitting on data your users can't reach?

The application layer is what turns a dataset into a product.