Built out of necessity, sold out of conviction
StockQL started the way most durable tools do: as infrastructure its builders needed and couldn't buy.
We wanted what every systematic researcher wants — a long-horizon, worldwide, survivorship-bias-free equities database we could query without limits. The vendors sell pieces of that: an API here, a bulk file there, each with its own quiet failure modes. Nobody sells the finished thing: a coherent database on your own machine.
So we built it, and we've run it in production ever since — daily ingests across 72 exchange calendars, tens of thousands of symbols, with an accumulating suite of validators written each time a feed surprised us. The partial bulk snapshot that looks complete. The exchange whose vendor metadata names the wrong country. The delisted symbol that wasn't. Each one cost us a debugging session so it doesn't have to cost you one.
StockQL the product is that system, packaged. We keep running the same code on our own research database every day — which means when a provider changes something, we notice before you do.
Questions, skepticism, edge cases we haven't thought of: hello@stockql.ai.