Our Platform Is Open For Business
Three years ago, when Mike and I founded Ozlo, we started building an index of knowledge about the world. A lot of people thought that was crazy. Google already has an index, why build another one? Our bet was that an intelligent future of voice interfaces and AI powered systems would need a different kind of index to be truly transformational. To deliver on the promise of AI and create new market opportunities and delightful, engaging user experiences, these systems would need to be backed by an understanding and reasoning system that didn’t exist yet.
Today, we’re excited to announce the launch of our platform. Intelligent systems powered by Ozlo are able to have longer, more meaningful conversations that convert more frequently and increase customer engagement. Ozlo’s knowledge index is the first to offer probabilistic assertions alongside accepted facts — creating a fundamentally better model that reflects the nuances of everyday life.
The index forms the foundational layer of our platform, which is a powerful suite of APIs that can be used to build an intelligent assistant from the ground up, increase the knowledge and understanding of an existing assistant technology, or power any system requiring deeper knowledge.
Our full stack approach allows companies to integrate at any level, leveraging our intent management and conversational systems, as well. Our products, Data, Intent & Converse, are available separately or as an integrated solution.
The Holy Grail — an intelligent interface that knows what you mean and responds accordingly — isn’t what we have in our kitchens, cars and living rooms today, and it’s not what we’re carrying around on our phones, either. 2016 may have been a breakthrough year for voice assistants and other “intelligent” platforms, but we’ve yet to see a truly delightful experience — the assistants of today show their brittle edges quickly. The Economist summed up the current state of Conversational Interfaces nicely:
“Open the pod bay doors, Hal.”
“I have a list of eBay results about pod doors, Dave.” ¹
Our goal is to offer solutions that quickly recapture the vast amounts of customer intent that are lost to these brittle edges and frequent failures and enable companies to fulfill on it for their customers.
Why the facts aren’t enough
What is a probabilistic assertion alongside an accepted fact? Great question. And, as it turns out, an important one. For much of the information we seek on a daily basis isn’t pure fact — it might be subjective “is the coffee good at Tartine?”, temporal “Is the skating rink at Squaw Valley open” (in general, yes, but not in 2017, when it is closed for repairs) or controversial “Is Delfina good for groups?” (sources differ).
We’ve built a system that ingests information, lots of it, extracts meaning and conceptual relationships and then constructs an ever-learning real world model, one that is intelligent in the way that humans are. It understands that a restaurant can be both noisy and quiet, that Philz both serves an Old Fashioned and doesn’t serve cocktails (their Old Fashioned contains coffee, but no alcohol). Reasoning precisely in an area where the facts are imprecise, Ozlo’s model of the world enables practical, applied AI solutions to real-world problems.
Ozlo’s knowledge increases over time. We’re pleased to announce today that we’ve increased the scope of our index to include recipes, movies & television, in addition to our existing knowledge of over 20M US places. We’re aggressively adding domains, such as general knowledge, to increase the utility of our index
Imagining and building toward the intelligent future, where assistants can carry on meaningful conversations and websites and apps can better anticipate customer needs requires myriad leaps of faith and the willingness to count on technical breakthroughs that are within a reasonable risk tolerance. That’s what we’ve done, and today we’re delivering. We couldn’t be more excited to share it with you.