Matthew Zeiler wanted to solve a simple problem: The mess our photo files have become.
For example, you're really excited about a selfie you took with that famous person you met while walking down the street, was that yesterday or the day before, or last week, or ... wait, what were you looking for again?
"You start scrolling through your phone collection, you get lost, that whole excitement gets lost," Zeiler said.
There is actually a simple solution - go through all your photos, one by one, and tag them.
Because that's gonna happen.
Zeiler's company, Clarifai, is built on machine learning that is able to "read" images and categorize them according to objects, emotions, events, places, and people, among other aspects. He launched Clarifai in 2013 with an API that enabled large enterprises to use the technology for different projects. Video-recognition tech was added in January.
Vimeo, for example, uses Clarifai's technology to properly tag videos so they perform better in search. Users upload videos all the time, but may not tag them in all the ways people would be searching for them. Clarifai can recognize mountains and sunsets and people, and all sorts of other items, neatly tagging the videos.
A medical company is using the technology to automatically characterize certain types of infections, bruises, and other health issues. This enables the doctor to immediately hone in on certain areas of concern, making her much more efficient. The company provided a large library of medical images to help train the Clarifai engine. The tech can learn anything - it just needs a glossary to initially interpret what it's seeing.
So about that whole B2C issue?
"Almost every investor said we couldn't do both," Zeiler told me this week. "Said we couldn't do both consumer and enterprise. We set out to prove them wrong."
It helped that Zeiler already knew he had a market for it: It helped him organize his own iPhone photo collection, and everyone he talked to had the same problem. So he built the company on the B2B, enterprise API. Other clients include Curalate, Unilever, Kodak and Groupon, all using the tech for different purposes, seamlessly under the hood.
Today, Clarifai released their first consumer-facing app, Forevery, in Apple's App Store. I took it for a spin to see if it really could make sense of my mess of a photo roll.
I allowed it to access my location and synced it with my address book so it could pull in names of people to attach to photos. All I had to do was tag one photo with a name and the app offered me a slew of other images that might be the same person. It was pretty accurate.
I have two sons, and tagged one photo with one son's name. It automatically pulled in several photos of him into one folder, and offered me a few others it thought were him as well. I did the same for my other son, and it automatically pulled in all the photos that had both my boys in it.
For pet lovers, you can pull all the images of your pet in, and Forevery will learn that your dog is not just a dog, but is your dog and will put future photos of him in that folder. For sharing, it learns who you share certain images with, and if you always share photos of your children with your spouse, it will recommend that option to you when you take photos of your children.
It will not automatically share any images with anyone else, and you don't have to use that aspect.
As I put the app through its paces, I desperately wanted to sic it on my iPhoto library and then take it into the basement, point it at my boxes of photos and say, "Organize!", as if it were some magic device.
The bad news is that it's only on iOS for now. The good news is that an Android app is coming. Desktop, web and other versions could be in the offing, depending on what users want from it.
All those investors who told Zeiler he couldn't build a business that served enterprise as well as consumers never intimidated him. He knew both sides could benefit the other, with consumers finding use cases that could work for enterprise, and with businesses using the technology in ways that would make the consumer product even better.
Plus, he thought of Foursquare, a company that started out consumer-facing and built a very strong technology and database that now is used via API by many businesses.
So what did all that naysaying give Clarifai?
"I would call it a fire," he said. "Having that fire in your belly."