Dumbfounded: A chat with Chris Seline, CEO of Dumbfind

What is Dumbfind? What was it? And what do you hope it will become?

Dumbfind was a search engine. Dumbfind is now a search engine with a twist! Dumbfind will become a suite of creative and different tools to find information in new ways. There are many ways to manipulate data to discover new knowledge, traditional search engines have barely tapped the potential of the information they have in their indexes. There are many very cool, innovative features coming to Dumbfind that leverage this information. And that’s before we get the users involved in the process! I am VERY excited.

The market seems to be overrun with search engines. Why launch a new one now?

There are a lot of search engines, but precious few have anything new to offer the world. There are only about 12 or so, at my last count, search engines that serve their own index. Most are meta searches, combining the results of several search engines together. Very few add value on top of that. One of those is Clusty. Clusty has proven there is room for a new search engine if you offer something innovative and useful. We would like to go way, way beyond Clusty and offer a wide array of tools that are just as useful as topic clustering. I honestly love this stuff, and have developed many tools over the years for manipulating web data. I want to open up a lot of those tools to the world so that everyone can benefit.

How much time do you devote to its growth? Do you have a day job?

I try to work as much as possible! And this is my day job! I quit the real working world in November of 2003 to develop Dumbfind full time, then I brought on a private investor in June of 2005 to raise the capital to hire some developers, buy decent servers (had 30+ homegrown servers in my attic), and move into office space with a big fat internet pipe. Before I raised the capital, I would work about 120 hours a week for months on end out of my house. Part of me misses it, but the other part of me actually likes being sane.

How large is your team and what are your backgrounds?

The team is still very small, just 5 people. I am being VERY careful about who I hire. This kind of work isn’t for everyone, if you don’t love data then you probably won’t make a good search engineer. You need to think like data. Most developers just think differently. I actually have no significant formal training in computer science, but I hire people that do.

 

 

What technologies are you currently using?

Java! On Linux! A very odd combination of jvm’s for different tasks on a few different Fedora cores to be specific. Java has some faults, which is why we use different jvm’s for different tasks, but you can’t beat its combination of flexibility and speed. I wrote an appserver in about 3 hours that we still use on the backend. I can’t imagine trying that in another language. In fact, it was serving the last beta site directly, but we threw Tomcat in front of it for the current release.

What is the greatest challenge to your success?

Getting people to care! You could make something twice as good as Google, but most people won’t bother to try it. Not that I am saying we are better than Google. We are offering new ways to search, some searches on Dumbfind are hard or impossible to reproduce on Google. Literally about half the people that try Dumbfind just search for themselves and then see that Google has more results. They are more comprehensive. No question. By a factor of about 15. The fact that we even come up with good results for most searches is a pretty good testament to the algorithms we are using. They still need tweaking, but just expanding the index can account for a lot of the difference between Dumbfind and Google. Reinforcement of links in the index is a huge part of determining relevancy for search results, so size does matter, if indirectly. We just can’t afford to expand our index at this point. Hopefully soon.

Do you have a business model? If so, what is it?

Yes! Advertising! And the model is FAR from fully explored. Google only serves ads on about half of their search results, by their estimates. There is plenty of opportunity to serve very relevant ads on roughly 90% of the other half. We have a new product coming out, hopefully end of February, that will help fill this gap.

What is the one thing you’re most proud of about the project?

That is hard to say. I feel like every part of it needs improvement before I can be proud of any of it.

How many hours of sleep do you get a night?

Roughly 0-9 hours per night!

What have you not told me about Dumbfind?

Despite my quite long-winded answers, I don’t feel as if I have told you anything! I am not a very talkative person by nature, but I could talk for a week on this subject and still only tap the surface of what we have planned. Thanks!

Thank you Chris!