I’m extremely happy to be able, finally, to talk about a product we’ve been quietly working on for more than two years: Ozlo.
I introduced Ozlo’s co-founders Charles Jolley and Mike Hanson to each other in 2013 when Mike was an Entrepreneur-in-Residence here at Greylock — coincidentally, the same year that mobile internet usage exceeded the PC.
What they noticed then, to paraphrase Louis CK, was that “…everything was amazing, and everybody was unhappy.”
Everything was amazing because the combination of smartphones and the global cloud meant that we had access to just about any information in the world, all the time.
But everyone was unhappy because we were still mostly searching for it like it was 1999.
We were typing tons of keywords into search boxes. Still sifting through lists of blue links. Still trying to find information embedded inside web documents instead of accessing it directly. As a result, we weren’t asking our phones some of the most basic & simplest questions — we’d all been trained only to ask things that we thought might be “Googlable.”
Mobile is different, and the interactions and questions and answers should be different, too. Our phones know a lot about where we are and what we’re doing. We’re only paying partial attention to them — we’re out in the world and distracted, not focused like with PCs. We mostly use our phones to communicate — not at all like traditional search engines. Above all, when we turn to our phones with questions, we want answers fast.
It seemed clear that there was an opportunity to build something more modern — higher signal for our mobile lives, much faster to get to information, but mostly just better able to answer the questions we actually have when we are out in the world, living our lives.
So Charles & Mike got started. They built Ozlo, which is launching today as an iOS app but will eventually have many more ways to interact with him. Ozlo lives on your phone and answers questions like, “Where can I find good coffee within walking distance?” Traditional search engines would struggle with a question like this. That’s a point that’s worth emphasizing — even in 2016, traditional search approaches aren’t great at answering basic human questions.
Here’s a post from Team Ozlo that shows you a a few things about how he works and what it can do today. At the moment, Ozlo knows about restaurants & foods — over time he’ll learn about more domains.
I’ve been using Ozlo for some time now, and watching him learn and improve nearly every day. I use him because he offers genuinely useful information just about every time, and because he does so very quickly — generally with just a tap or two, and often in less than a second. I haven’t really found anything else quite like it (and believe me, I’ve looked.)
I think it’s unique partly due to the design and partly due to the deep technology stack that Team Ozlo has built. The design stands out for me as the best example of a conversational interface that I can find. Ozlo’s not a command line bot; it’s deeper and richer. He’s also not pure text — or rather, he can be sometimes, but also will use more sophisticated UI elements when he’s able, like in this native app. Ozlo gets the back & forth exploration of a real conversation. He knows things about who I am and what I like. And his interactions are fast & fluid in a way that feels natural and native.
The technology stands out because the team has blended approaches from the worlds of search, artificial intelligence, natural language processing, and even interactive fiction — it’s a full stack developed to build rich conversational experiences, and it shows.
We know that we’re still on the very early part of this journey — lots more to build, and lots more to teach Ozlo about over the coming months and years. So consider today’s product a down payment on the future — a system that’s useful now and will be more useful tomorrow & the next day & the next one after that.
They’re opening up a few thousand slots today on the beta list, so if you’re interested in using Ozlo and helping it learn, and being on the journey, head over to sign up at Ozlo.com.
On a very personal level, I’m extremely proud of Team Ozlo. Charles and Mike first set up shop in the Greylock offices and proceeded to build a first rate team of search, AI, language and design builders — no surprise since both Charles and Mike are exceptional in their ability to innovate and build complete systems from data & algorithms through to new types of user interfaces. I’ve known both Charles and Mike for years, and really love working with them and the Ozlo team.
Ozlo provides something new: a quick & easy way to have a conversation about the world around you, supported by an extremely deep technology stack. It’s hard to overstate how difficult it is to make something so challenging feel so simple, and it’s great to finally be able to share it with the world. (Or maybe not the whole world just yet, but, you know, lucky beta users — go sign up! 👊)
John's Blog 