With each passing month, Repustate has grown in popularity. More and more businesses rely on Repustate to analyze the data that flows through their systems. A core tenant of Repustate has been that we never wanted to throttle our API if we didn’t have to; we wanted our users, particularly our paying customers, to be able to push the limits with our sentiment analysis API.

This all came to a head this past week. Many of you might have noticed intermittent down periods over the past week. While our feature development has kept pace with our customers’ demands, our infrastructure and ability to support the ever-increasing load has not. Over this weekend, we took several steps to mitigate these issues.

We greatly expanded the number of servers we use and the capacity of each server. As an aside, I can’t say enough about the good people at Liquid Web and their Storm on Demand offering. If you’re looking for SSD servers, use these guys, they’re awesome. We made great improvements to the software that helps run Repustate, implementing many performance improvements and optimizations, which deserve a blog posting in their own right.

Early signs show that all of these changes together have resolved the problems we saw last week, however, as always, please contact us if you have any questions ro issues.

This is also a good time to announce that we’re releasing a Repustate appliance that you’ll be able to deploy locally on your own server infrastructure. More details to come shortly, but long story short, it’ll have all the goodness of Repustate’s API, but with the added benefit that you’ll be able to tune things the way you like and you’ll have absolutely no API usage limitations.

Finally, a big thanks to all of our customers who have been patient with us during the past few days.