A new Seattle-based artificial intelligence startup called KITT.AI, incubated inside Paul Allen’s Allen Institute for Artificial Intelligence (AI2), has landed funding from Founders’ Co-op and Amazon’s Alexa Fund — preparing to launch a platform that will let developers build apps, devices and services that communicate conversationally with users through natural language.
The company was co-founded by Xuchen Yao, a Johns Hopkins University PhD graduate who last year joined the AI2 incubator, led by Oren Etzioni, the former University of Washington computer science professor who now heads the Allen Institute for Artificial Intelligence.
The startup’s other co-founders are Guoguo Chen, a deep learning and speech recognition expert who created the “OK Google” hotword detection prototype for Android; and Kenji Sagae former professor of Natural Language Processing at USC and an expert in natural language parsing and dialogue systems, as noted by Chris DeVore, the Founders’ Co-op co-founder and general partner, in a post announcing the funding.
KITT.AI has already developed a prototype called Semantic Lighting based on its technology — letting people interact with smartlights by talking with them — demonstrating the early capabilities of its natural language processing platform. The team behind Amazon’s Echo smart speaker was intrigued by an early demo, and the company’s $100 million Alexa Fund ended up participating in the startup’s financing round.
The companies aren’t disclosing the size of the funding. KITT.AI is planning to build an “Alexa Skill” based on Semantic Lighting, extending the capabilities of Amazon’s virtual assistant so that people can not only turn lights on and off, but also do colorful things with lighting through speech.
KITT.AI’s bigger push is building an open platform and community for developers, providing natural and spoken language understanding as a service. The startup plans to go to market this spring with two products: one in speech understanding, and one in text understanding, Yao explained via email this morning. The service will be free to individual developers, with a paid version for enterprise customers.
DeVore said via phone that KITT.AI promises to do for conversational natural language processing what Twilio did for telephony, creating a platform that anyone can use to build the technology into their products.
“Think of this as the beginnings of a generic platform for anyone who wants to have voice input or voice control for their application, but doesn’t have the scientific wherewithal to do that,” DeVore said. The company is “creating a developer-facing platform that allows any developer to put a conversational voice control layer on their software.”
KITT.AI is part of a larger trend of the Seattle region developing new strength in these areas, DeVore wrote on his site, “Between UW CSE, AI2, Amazon, Microsoft and Google, Seattle has quietly developed a very deep pool of engineering talent in artificial intelligence and machine learning, and it’s exciting to see entrepreneurial offshoots of that ecosystem begin to flower.”
With three research-oriented co-founders, the company is now “actively looking for our first couple of employees who will be in charge of developing and engineering,” Yao said. It’s looking in particular for people skilled in Django, Node.js, and cloud deployment.
Where does the name KITT.AI come from? No, it’s not meant to be a reference to the Knight Rider car, Yao said. Instead, he said, the origins were in the concept of an “AI toolkit,” with an extra “t” thrown in.