Data & analytics platforms

Why these products, and why this category

Data platforms sit at the centre of public-sector IT, which means the federal award data names them often enough to build a real evidence base. Snowflake, Databricks, Tableau, Informatica, Alteryx, Qlik, MongoDB and their peers appear in hundreds of awards across dozens of agencies, from the Department of Defense to the Department of Health and Human Services and the Treasury. That spread is what makes a page worth writing: a figure drawn from one agency tells you very little, but dozens of awards across dozens of agencies let a buyer see the shape of what is actually paid.

The category is not hard-coded. It is a value in config/categories.yaml, with its Product Service Codes, its alias file and its time window. The same pipeline runs for a second category without a code change; only the configuration differs.

How a product earns a page here

A product is published only when it clears a fixed threshold. It needs at least three distinct public awards — distinct award IDs, not line items — or one award together with a verified vendor list price. Every award must carry an award ID, a recipient, an awarding agency, an amount, an action date and a resolving USAspending URL. The product must be named explicitly in the award text, matched as a whole word, with a confidence above the configured minimum. A product that fails any part of that does not get a thinner page; it gets no page, and its reason is recorded in rejected.json.

The distinction that matters most on every page is between a contract total and a unit price. A seven-figure award for “software implementation services” is not the price of a licence, and this site never presents it as one. Each figure carries its type next to it, and types are only compared with their own kind. The full procedure is on the methodology page.

Products with evidence