SimpleGeo Launching 'Storage': A Distributed Hosted Database For Location Data

The story of SimpleGeo is a familiar one: two founders — Matt Galligan and Joe Stump — set off to create location-based games, only to find that the tools they wanted to use to build their apps didn’t exist yet. So they switched gears and decided to build what they wished they had: a suite of tools optimized for the creation of location-based services (which was probably a good call given the explosion of location-aware mobile devices).

The startup launched almost exactly one year ago, and now offers products including ‘Places’ (a database of POIs) and ‘Context’, which lets developers query for data relevant to a given location, like the local weather. And now, they’re ready for what Galligan calls the company’s “grand unveiling”, explaining that it’s what they’ve been focused on for the last year: SimpleGeo Storage, which will be going live tomorrow.

‘Storage’ is a new, hosted database that developers can use to store and quickly retrieve data that has location information associated with it. And Galligan says that it has some key optimizations that make it significantly faster than other available database solutions. During a month-long beta-test with multiple partners, SimpleGeo saw 99% of spatial queries with processing times of less than 100ms (and engineer Mike Malone says that average was more like 20ms).

So what makes SimpleGeo Storage different from other database solutions that have been used for location-related apps? First, there’s the fact that it’s hosted, so developers can outsource their infrastructure issues (all data is replicated in at least two datacenters). It’s also distributed using Cassandra, so there’s no single point of failure. Storage is actually hosted on AWS, so it’s a ‘cloud built on a cloud.

Galligan says that Storage also has special optimizations related to dealing with location — it uses a tree structure for managing data which allows for efficient queries. It isn’t the only database to do this — PostGIS, for example, uses R-trees — but Galligan points out that even if PostGIS could match SimpleGeo Storage’s sub-100ms processing times with ease, it doesn’t include the same pre-baked APIs, reliability, or scalability afforded by SimpleGeo.

As a hosted service SimpleGeo is charging based on usage (there aren’t any plans to open-source the database). Storage is 25 cents per thousand data calls, and 10 cents per thousand records stored.

Galligan is very optimistic about the future of the product: he points out that location information can be attached to a huge number of things, from transactions to virtual goods, to mobile phones. And he’s hoping that SimpleGeo becomes the definitive solution for managing this data. However, SimpleGeo isn’t alone in this question — other developer-focused location tools include Location Labs, and Microsoft’s Azure has some location tools as well.

[youtube http://www.youtube.com/watch?v=7J61pPG9j90&fs=1&hl=en_US]