The past couple years I’ve developed a growing interest in building and managing scalable systems. I actually only recently came to the conclusion that this is what’s been going on, but that’s pretty understandable considering I have 2 girls, a 3 year old and 11 month old. I think I’m now slowly coming out of that sleep depravation haze that parents of small children go through.

So during this time, I’ve been working on Unscatter.com. It’s gone through several iterations. I built it on Appengine at first, then moved it to cloud machines at Rackspace. It’s served the following functions, yes all built to workable prototypes.

1) Self service based search for businesses. At one point you could create an account on Unscatter.com and set up a search engine for your own business, or topics. You could set up a list of 10 domains to search, content coming from Yahoo! Boss v1. Then you could add Facebook, Twitter and Youtube accounts to also pull data in from. There was a Boss backed image search, site specific and a Youtube based video search. The example site I had set up was for conservation, pulling data from the websites and social media assets of organizations like The Nature Conservancy, Greenpeace and the World Wildlife Foundation.

2) The next implementation of Unscatter.com was a Twitter and Facebook client. It was pretty simple but functional. Sign in with Twitter and Facebook, get your feed and posting was supported. You could sign in with multiple Twitter accounts. It met my needs at the time.

3) Next I tried the news portal route. I used the Bing API to put together a basic news portal, then added in Fwix to get local news as well. I used the YQL-Geo library to manage getting a users location. A friend of mine is a journalist for a local paper in a small town in Virginia, and she used this to find news from her area and found it pretty useful.

4) I then went back to search. I was researching new cars as the SUV we owned wasn’t providing a good fit for our family. I’d start with Google, then check Blekko using /date and /reviews. Then I’d try Twitter to see if anything turned up there, basically trying to get an overall feel for peoples comfort with cars. At this time I was also interested in technologies like MongoDB and the Tornado web framework which I had been using for a while so I was always looking for the most recent information about them. It was a lot of bouncing around the internet and then I thought “umm, I do know how to use APIs and I do have unscatter.com as a domain”. So I built the current iteration of Unscatter.com. It gives me one place to search all the sites I was checking individually. I also added some nice to have features like the ability to bookmark multiple searches so that I could have a quick link to a daily search to perform. It just hit me while writing this blog post that maybe I should call the concept search bundles.

Now through all this, I also thought I wanted to be an entrepreneur. When our youngest was a few weeks old I was actually pushing myself to build a MVP and get it out, this was early in the current iteration of Unscatter.com. However, I quickly became frustrated with trying to get time to build it and made the decision to forget about trying to start a business right now. I decided spending the time with my family was more important than secluding myself away to build something, so I turned development on the site into a hobby. Productivity dropped a lot of course, but didn’t stall.

Recently, I’ve been diving in ZeroMQ, building a scalable infrastructure I plan to build on Unscatter.com on top of. I’ve been telling myself because Unscatter.com is a hobby, I don’t need to follow the whole process of building a MVP and scaling later. I’ve come up with a pretty large plan for what Unscatter.com will be in the process.

The reality is, I don’t think I want to be an entrepreneur. Well, I don’t want to be a CEO anyway. I want to build cool stuff. I want to build scalable infrastructure and manage it, really. My current position doesn’t give me the opportunity to do this, so I’ve basically created the challenges for myself by building Unscatter.com. At my current position I generally deal with COTS and manage software update processes. There’s not a lot of development work, primarily it’s just bash and the occasional python scripting.

I’m not going to stop working on Unscatter.com by any means. I’m still building a product for at least one user, me. It’s just interesting that for a long time I thought I was building a business, but in reality I’ve just been giving myself an excuse to learn things that I haven’t had a need to learn at my day job. I think I might just be addicted to learning.