On June 2005 I left BeInSync to start investigating new ideas and directions, and founded 3D3R Software Studios.

The first result of this work was WonderFrog, an application that, in real-time, detects Hebrew/English typing mistakes (like typing akuo instead of שלום) and corrects them automatically.

This was really great, really cool and really satisfying. The client was released, there was some buzz, and all nice, it still gets new users each day, and has some tens of thousands of users.

The problem with the client, was that it sat on the main line of communication between the user and the computer… the keyboard. This means that if the client is not 100% perfect
, and by not perfect i mean ‘miss out on a word’, the user will notice this, because users hit keys on the keyboard all the time. This in turn made users angry and left more to desire from adoption rate. Just a short notice: this kind of application can not be perfect, just like the MS-Word spell-checker can’t be 100% perfect, text and words are flexible, too flexible for a computer to always determine what exactly the users intention was.

That was a great lesson, it taught me not to lie to myself. I couldn’t use the client because it drove me crazy after some time, it did the same thing to most of the other users out there.

So the next step was just to figure out where we can use this that won’t drive users perfect. Search engines! lets use WonderFrog in search engines, we’ll see if they typed in the wrong language and suggest the alternate typing. This is fantastic, the mere distinction between forcing the application’s logic onto the user and between suggesting to him the correct typing makes all the difference. We do lose the real-time feature but we gain users, zillions of them, and their all happy because our app can only help them, if it makes a false suggestion they just ignore it, because their brain filters it out, and if our false-positive rate is low enough, we’ll do good.

So, we made a server version of WonderFrog, did a promotional video (5 minutes, watch it) for search engines, and I went on a hunt to catch Yahoo, Google and Microsoft Live. I reached the right people in these three companies, but nothing came out of it. Its too small of a technology for these kind of companies to purchase, that’s the best answer I got from the people I consulted with. I’m also new to pitching and selling, experience comes from experience and you’ve got to start somewhere.

The Israeli search engine Walla introduced this feature a month ago, and Google introduced it yesterday.

Who knows, maybe Google played cold with my video because they were already developing it. Maybe it gave them the idea, and maybe it just ended in Marissa Ann Mayer’s spam folder

It doesn’t even matter that I filed a provisional patent for WonderFrog on May 2006, there is too much of a black hole around prior art.

That’s my story, I enjoyed it, learned a lot from it, and have charged forward with new ventures.
I guess not hitting the jackpot is something essential on the way to the big gong.