Matt, Kyle and I are wrapping things up for the public launch. I have spent the last week writing Nagios alerts for some really esoteric details of our system. We're sending Persai out into the world in July, if everything goes right. It has been a long time coming - the system has been in development for over a year.That Ride Sucks, Anyway
Why is this thing taking so long to launch? We've had a lot of failures along the way, but we've finally converged on a system that I believe to be correct. The focus of the product has changed a little bit, too.
Our pitch used to be: a content filtering and delivery system that learns about what you're really interested in, and shows you news based on your feedback.
That doesn't sit too well with investors, it turns out.
Now, it's: a persistent search that learns the nuanced details of what you like.
That one picked up a few ears, and got a few pens working.
We ended up re-writing the internals of this beast three times before we had what's running today. Now it's lean and mean. Our classifiers went through a massive reduction in feature space: about six orders of magnitude. Some of our linguistics algorithms got ten times faster, and we are crawling thousands more sources.
The next couple of weeks are going to be very busy and very exciting. We will likely have a launch party, but we're not pissing away our investors' money on it. It'll be a barbecue in the back yard of the "office" with a bunch of beer. Maybe 111 Minna next time.
