ReportGrid has found an excellent niche and positioned itself as the leader of the pack when it comes to web-savvy. It provides data analytics as a service, which is not new in itself. The idea of providing this service with data submission through its ReportGrid API and analytics delivered in a web browser really takes it to the next level. ReportGrid is one of 43 analytics APIs, and one of 241 Enterprise APIs. Its simplicity makes it relevant to a broader audience than many of the more use-case tailored web services, such as those that focus on web site analytics or social media.
ReportGrid’s value proposition is that it will take your data through it’s incredibly simple API and provide hassle free and in-depth analytics, as if by magic. The analytics and visualization feature set is under active development. Automatic correlation detection is in the works, which sounds like a little bit of artificial intelligence if I had to take a wild guess. Just imagining that my data is sitting with a company that is devoted to mining it and periodically drawing out new correlations just makes me drool. A dream come true for a data geek!
I couldn’t fight my curiosity on this one. I just had to reach out to ReportGrid to find out more about this data crunching artificial intelligence. Here’s what founder John A De Goes revealed:
“In addition to basic correlation detection, we will soon be working on two additional features: anomaly detection and time series prediction, which are really two heads of the same coin. Both require machine learning.The first attempts to find irregular patterns in data, such as, for example, a spike of transactions that occurs every Monday mornings for certain segments of events. When you’re dealing with billions of events per months, sometimes the most important data is that which deviates from the normal patterns. That’s what anomaly detection is all about. Finding the needles in the haystacks.
The second attempts to predict trends in time series data. If you’re using ReportGrid to track pushes to mobile phones, for example, then you could use time series prediction to predict what’s going to happen next Thursday if you decide to push an event to mobile phones at 9:00 AM for everyone in New York City.”
Since ReportGrid is so web savvy, of course it’s API is RESTful and supports JSON. In fact, you can just submit JSON objects directly to it for storage. No re-formatting necessary! So far the API is quite clean and straightforward. Essential items are already there, such as a token based permissioning system for controlling who can access which parts of your data set. Querying via the API is also supported, which should make it easy to play nice with other components.
To top it all off, ReportGrid offers a very attractive pricing model which allows for intensive API usage for a reasonable cost. Is 2011 the year that your company will give outsourced analytics a try?