Lawrenceville tech startup's app lets fingers do the resting

| Monday, March 7, 2016, 11:00 p.m.

Chris Maury wants to have better conversations with his mobile devices.

Google searches, buying clothes online, looking up recipes — he doesn't want to hunt and peck on a tiny touch screen. Tell his tablet or phone to find a Chinese restaurant, and they should be able not only to find one, but answer questions about the menu and help him order.

“More of the devices that we're going to be interacting with, it's not going to have a mouse and touch screen,” said Maury, the 29-year-old CEO of Lawrenceville-based Conversant Labs, a tech startup that is developing voice-enabled app technology. “Voice is going to be the way we interact with these devices.”

Conversant has an app called Say Shopping that allows users to search, browse and buy items without typing a word. Target is the first major retailer to allow its goods to be sold through Say Shopping via an affiliation agreement. Target pays Conversant 2 percent of each order placed using the app, though Maury declined to say how much revenue it had collected from Target so far.

But Say Shopping is a test for Conversant to show its voice-enabled technology can work. The larger aim is to sell a software platform called SayKit to other app developers. Conversant will license SayKit to app developers and charge them a royalty of a few cents for every time someone uses it.

SayKit is in limited release now and is expected to hit the wider market next month.

The market for mobile apps will reach $73 billion by 2019, according to information technology research firm Garnter Inc., and voice-enabled technology is increasingly important to how apps work. By 2018, nearly a third of our interactions with technology will be through “conversations” with smart machines, Gartner said.

Voice-enabled apps are on the market now, but they have limits. Ask Google a question about the weather forecast in Pittsburgh, and it will give you the temperature. But it can't answer more advanced questions that can help you buy a raincoat.

“The human voice is not as logical as computers would like it to be,” said Werner Goertz, research director of personal technologies at Gartner. “So it takes a lot of artificial intelligence, a lot of compute power to understand when a human is being ironic, ambivalent. When a human is lying.”

Conversant is helping computers understand more complex contexts. It allows developers of shopping apps, for example, to create technology that cannot just read a screen or fulfill basic commands (e.g. “I want to buy a coffee”) but also respond with feedback (“Would you like a muffin with your coffee?”).

Ryan Strunk, 32, has used the Say Shopping app to buy items at Target, where he works as a senior accessibility analyst. Strunk, who is blind, met Maury at a conference on technology and disability in California and decided to give the app a try. Strunk found it a refreshing change from the screen readers he used.

“With Say Shopping, you can do an end-to-end shopping experience in less than two minutes,” said Strunk, who was speaking personally and not on behalf of Target. “That's amazing.”

Maury has a personal as well as professional stake in the company's success. He is losing his eyesight.

In 2010, a visit to his optometrist led to a diagnosis of the degenerative eye condition known as Stargardt's disease. He was a 24-year-old tech entrepreneur in Palo Alto who lived his life through screens. As his vision got worse, he discovered the only technology available to help him, called a screen reader, was both crude and in some cases expensive, costing several thousand dollars.

Even a basic Google search can be painfully long using screen readers on the market today. So, in 2014, Maury started Conversant to develop better tools. Last year, Conversant won Bank of New York Mellon's “UpPrize” competition for companies working to address social challenges. The award came with a $200,000 investment and another $200,000 in grants. Conversant also has backing from the business accelerator Innovation Works.

Widespread application

Conversant isn't limiting its market to apps for 7.3 million visually impaired Americans, Maury said. Its software is relevant to those who wants to use their devices hands-free.

But Conversant has some stiff competition. Online retail giant Amazon is trying to do something similar. Its Alexa software is part of Echo, a cylindrical device that can do everything from play music to help you find a Chinese restaurant, all by talking with it. And like Conversant, Amazon is selling an Alexa tool kit to third party app developers.

Separating itself from Amazon is Conversant's biggest challenge, Goertz said.

“There will be a plethora of third party vendors that license Amazon's (technology),” he said.

Maury gives Amazon credit for developing Alexa and selling it to a broader market of app companies, but he said Alexa is still limited in what it can do.

“You can ask a question and receive an answer or an answer with clarification,” Maury said. “Doing anything more complicated with more back and forth is not supported.”

A step ahead

Say Shopping has its bugs, as well, and still doesn't feel like a natural conversation, Strunk said. But it understands much more than other apps on the market.

“It's smart enough to know what you want based on wider set of key words,” he said.

Another challenge is getting consumers to adopt voice-enabled apps for wider use beyond basic searches such as the weather forecast or restaurants, Maury said.

This is the way so many technological innovations begin, said Jeffrey Bigham, an associate professor at Carnegie Mellon University's Human-Computer Interaction Institute. He has seen a demo of Say Shopping and says it is “likely a harbinger of the really great innovation we'll see in conversational interfaces going forward.”

“People with disabilities are often leaders in developing and using technology that we all often end up using,” Bigham said. “It's happened with the telephone, the (optical character reader), the acoustic modem, etc., in the past, and speech-based interfaces are just the latest of this trend.”

Chris Fleisher is a Tribune-Review staff writer. Reach him at 412-320-7854 or cfleisher@tribweb.com.

Subscribe today! Click here for our subscription offers.

 

 

 
Show commenting policy
Trending