Startups, Health IT, Hospitals

How Syllable’s AI chatbot is helping Seattle Reproductive Medicine answer patients’ questions

The Silicon Valley startup built a bot for SRM’s website that can answer patients’ questions about becoming an egg donor.

AI, artificial intelligence, chatbot, bot

As artificial intelligence chatbots are gaining momentum, one startup out of Silicon Valley is harnessing the technology to influence healthcare.

Syllable was founded about a year and a half ago by Kobus Jooste (CEO), Gui Bastos (COO) and Andrew Rogers. In early 2017, it raised a seed round from investors like NEA and Greylock Partners, according to Crunchbase.

The company’s conversational AI chatbot can be utilized in a variety of ways.

In a recent phone interview, Bastos discussed Syllable’s work with Seattle Reproductive Medicine, which has five clinics as well as two satellite locations. Bastos explained that the center encounters more women interested in becoming egg donors than it has staff members to address their needs.

To help, Syllable built a bot for the center’s website that can answer patients’ questions about becoming an egg donor. The chatbot shows up on the site as a purple and white voice bubble in the corner of the page. Users click on the icon and can type in queries like “How much do I get paid?” and “Do I have to live in Seattle?” The content for all the answers comes from Seattle Reproductive Medicine.

Still, the bot doesn’t know everything. “Sometimes we have answers, sometimes we don’t have answers,” Bastos said.

When it isn’t fully confident it has the right response, the bot poses other questions that might be useful to the patient.

In the case of “unlinked questions,” or queries the chatbot doesn’t have an answer to, SRM providers can use the backend of the system and add the appropriate answer.

The technology launched at SRM with about 100 questions in the database that the bot could answer. Within a few months, that number increased to about 1,000, Bastos said. He added that the more patients use and interact with the bot, the more it will learn.

But it is important to note that Syllable does supervised learning. “We don’t let the computer train itself,” Bastos said. “All the learning that the AI system does is supervised.”

In addition to its work with SRM, the Sunnyvale, California startup is using its technology to assist NewYork-Presbyterian with patient scheduling. “We realize we’re not going to replace an existing scheduling system for a large hospital, but we can definitely make it easier for [patients] to schedule,” Bastos said.

As a fairly new company, Syllable has experimented by bringing its know-how to a number of situations. For now, one of its key goals is to move its product out to more individuals and gain new customers.

Photo: Jull1491, Getty Images