U-M Startup Clinc is on a Mission to Democratize A.I. Technologies
After building Lucida, an open-source intelligent personal assistant, University of Michigan spinout Clinc has a new $225,000 grant from the National Science Foundation to help get its artificial intelligence technologies to market.
That’s on top of closing a $1.2 million seed round disclosed in March. The equity funding round was led by Ann Arbor’s eLab Ventures, with participation from Hyde Park Venture Partners and Cahoots Holdings. Clinc co-founder and CEO Jason Mars said the company is on a mission to lead a transformative shift where state-of-the-art A.I. research emerges from the lab to be more seamlessly applied toward solving problems in the real world.
Mars said he always wanted to start an artificial intelligence company, and his path to entrepreneurship opened last year after he and his Clinc co-founder, Lingjia Tang, created Sirius, an open-source virtual personal assistant.
“We found the algorithms used by Google and Apple, so we stitched them together to create an end-to-end open-source platform,” Mars said. “It made a huge splash, and from the moment we released it, we heard from large companies all over the world asking how to commercialize the technology for their products. There was a clear void in the market if you weren’t Apple or Google, and we decided it was a world need we could deliver on.”
After releasing Sirius in March of 2015, Clinc officially launched four months later. Despite the interest in Sirius pouring in from around the globe, there was one entity that took exception.
“We got a letter from Apple,” Mars explained. “They said they loved the project but not the name, because it was a little too close to Siri. We fought back for a while, but we decided it would be easier to change the name. So, in November, Sirius became Lucida.”
After the name change, the Clinc team continued to work around the clock to further develop Lucida, Mars said. “We want Lucida to be the open-source counterpart to Siri or [Microsoft’s] Cortana. The goal is to build artificial intelligence that solves problems in people’s lives.”
According to Mars, what Lucida offers is a platform for cloud A.I. applications. The federal grant will help Clinc improve the intelligent capabilities of Lucida and build a scalable technology infrastructure to power the digital engines required to run deep learning programs. Lucida is also designed to allow developers and the open-source community to create and deploy custom intelligent assistants in ways that are beneficial to society, such as improving technological access for disabled people.
Mars said Clinc’s technology has the ability to transcend software code’s traditional rule-based user commands and canned answers. Clinc instead focuses on natural language processing to mimic human intelligence and communication. Because the company’s platform automatically learns over time, it’s not dependent on continual coding to grow, he said.
Mars said the 11-person Clinc team will make money building its customers personalized engines that “live in the Lucida platform” and selling them as a subscription software service. He was especially excited, he said, about “democratizing” the kinds of artificial intelligence technologies used by Microsoft, Google, Apple, and other tech giants.
“Understanding our own data is one of the problems we’re trying to tackle,” he said, using a person’s accumulated financial data as an example. Most people don’t have the time or inclination to pore over months of bank and credit card statements looking for useful patterns and information, he said. “As new data is generated over a lifetime, we have a very poor understanding of that story. But we can build an intelligent assistant to understand requests and provide insights.”
Mars declined to go into much detail, but he said Clinc plans to release a financial technology assistant in September that will create instant insights on spending. He said the industry’s reaction to Clinc’s technology so far has been “super exciting.”
“A lot of people are trying to solve this problem using a top-down approach,” Mars added. “But what we’re doing is creating an experience with completely natural, unbounded speech.”
Meanwhile, Mars is also hard at work on the faculty team behind a $4.5 million collaboration between IBM and U-M called Project Sapphire, which is creating a virtual academic advisor that will be built on top of Lucida.
“It’s a really nice problem—we have a lot of unstructured data and a host of student data,” he said. “We’re building the A.I. to extract insights and create a natural language experience. IBM is crazy about this project because in A.I., data is gold. Getting access to build this kind of conversation is what drove them to invest in the project.”
As far as Clinc goes, Mars is excited to see what happens when technologies he’s spent a good chunk of time perfecting in the lab are released to a wider audience.
“It’s a team effort—an awesome family of researchers working together for years and coming together to build a company,” he said. “I’m having the time of my life.”