Data is commodity. Computation is the product.

There's a version of this post that's about proprietary data partnerships, exclusive feeds, and moats. This isn't that post, because that isn't how this API is built.

Airport records, weather observations, and reference data for navaids and airlines are, at their core, widely available to anyone building in this space — the barrier was never possessing them. Live ATIS broadcasts are the one piece that requires an active feed rather than a static dataset, but even those broadcasts are public over the airwaves — any pilot can tune the frequency and hear the same thing we're parsing.

None of this is a secret. Anyone building an aviation product has access to similar raw inputs.

So what are you actually paying for?

Two things, and neither of them is data exclusivity.

The integration work. Different domains mean different formats, different update cadences, different sets of edge cases. Turning all of that into one consistent schema, reachable through one API key, is genuinely time-consuming — that's the work we've done so you don't have to.

What we compute on top. This is the part that actually differentiates us: active runway determination from wind data, structured ATIS parsing instead of raw teletype text, per-runway surface conditions extracted from the same broadcasts most tools discard after reading the wind line. That's not raw data repackaged — it's a computed answer that doesn't exist anywhere upstream.

Why this approach, deliberately

We could try to lock down exclusive data partnerships and sell access to information you can't get elsewhere. Some aviation data companies do exactly that, for the pieces of the puzzle where it's actually possible — global ADS-B coverage, for instance, requires real infrastructure, which is why it's on our roadmap rather than shipped today.

But for airport operations data specifically, the raw inputs are widely available, and that's not going to change. Betting the product on access to information that isn't scarce would be a bad bet. Betting it on turning that information into something more useful than its raw form — that's a bet we can keep winning by doing better work, not by controlling a pipe.