[shahary]

Thanks! What we can share is that we have essentially three main parts - a custom NLP engine that is built specifically for texting (consists of numerous phases of filtering and normalizing), this is the core of Guggy and will be also used in the future for more entertaining applications.

The second part is the creatively tagged image repository. Even though tech enables now to tag objects in images, it still can't do it in a creative and funny way, so for the moment we were aiming for quality rather than quantity (I do agree that massive existing GIF repositories do not yield desired results), whereas we have plans for extending our repository significantly in the future.

The third part is also what we consider a big advantage of Guggy, and that's the image rendering engine. That allows us to personalized each request, such that different users could potentially see different results for the same texts (imagine seeing them in a font that you like, and your friend sees them in a different font, but there are also other applications for this). The images are therefore rendered on the fly per each request and are never stored.

Hope that clears up the questions you had in mind.

[dom96]

Thank you for the reply! It certainly does.

Something that you guys might find interesting. Even after a day I already forgot exactly the name of Guggy, instead attempting to navigate to http://guppy.com and http://gubby.com. Probably just my bad memory, but might help you guys out if you see bad retention.

[rotemy]

That's interesting.

We will think how to make our brand more memorable, thanks.