Hello HN,

If you geocode a lot of addresses (logistics, real estate, insurance, etc.), the cost can become significant because most pipelines rely on a single commercial provider.

While working with large datasets, I noticed many addresses could be resolved correctly using cheaper or open providers first.

For example: OpenStreetMap -> HERE -> Google

Only falling back to the more expensive provider when necessary.

In some datasets this reduced geocoding costs by 60–90% without reducing match quality.

So I built a small tool to experiment with this approach more easily. It lets you:

- define a cascading order of providers

- set acceptance rules (e.g. minimum confidence)

- analyze success rates and failures

The reason why I created this is because I noticed big players (I mean, people geocoding a lot) are already implementing such things (but have a hard time maintaining it), so it could be useful for "small"/"medium" size players.

I'm curious how others handle geocoding pipelines at scale. Do you rely on a single provider, or do you cascade multiple ones?

What do you think about this project ?

Thanks, F.