When the co-founders of Appurify sold their app development company to Google and started working for the search giant in 2014, they were impressed with the company's culture. They were also surprised to find that Google faced a lot of the same problems as other companies when it came to onboarding new employees.

"They obviously have a good internal search feature," says co-founder Jay Srinivasan, "but we still spent so much time looking for information, so much time looking for services. We saw firsthand how hard it was for new employees to ramp up."

Cue the aha! moment for Srinivasan and fellow co-founders David Kaneda and Pratyus Patnaik. After two years at Google, the trio set off to start their own venture.

The result is Spoke, a startup that uses artificial intelligence to try to make employees' lives easier. The San Francisco-based company's software learns information about the company and answers workers' inquiries across various channels via a chatbot. Launching in March, Spoke's platform promises to make life easier for HR departments, IT workers, and office managers everywhere. It's the newest entrant into a field of workplace chatbots, a budding corner of A.I. that will soon count Slack among its competitors.

"The approach we wanted to take," Srinivasan says, "was, how do we build a platform or tool that makes it easier for organizations to manage and respond to internal workplace requests?" For new employees, that could mean a searchable home for tax forms and employee handbooks. For the whole staff, it could mean a bot that knows everything from the guest Wi-Fi password to how to input travel expenses to the company policy on rolling over sick days from year to year.

When companies first install Spoke, they can program it with the answers to the dozen or so questions that employees ask most often. For everything else, Spoke learns on the go, gaining knowledge and refining its answers whenever workers indicate that an interaction was successful.

The software is designed to be more than just a source of information; it also can handle facilities requests, such as meeting room and equipment reservations, and can pass requests on to office managers when necessary. Some of the clients in Spoke's pilot program are using the tool as a ticketing system to prioritize IT requests.

Srinivasan says that 100 companies are part of the program, ranging from nonprofits to agricultural firms to tech startups. One client, marketing firm Digital Pi, has no central office, so its entirely remote workforce has been using the platform to find the materials they need. Another, motorized scooter rental company Scoot, deployed the software to its mechanics to provide them with quick answers while they're out in the field.

That's where Srinivasan sees a particular advantage. You can use Spoke within its cleanly designed desktop and mobile apps, but you can also email it, text it, add it to Slack--or, eventually, talk to it via Alexa. "Wherever you are," Srinivasan says, "Spoke comes to you."

Spoke is not, however, a source of general knowledge. The app won't find pizzerias in your area or suggest a song you might like. 

Srinivasan says that's by design. "A.I. tends to work much better if you sort of narrow the focus of how you use it," he says. "Spoke is not this magical bot that can answer any question you have about anything. But it's very good at answering those questions that get asked over and over again so your employees can get back to things that matter." Sticking to what Spoke does best--and setting customers' expectations accordingly--will be key to the startup's success. 

Equally important will be overcoming what soon might be some stiff competition. Slack is working on its own A.I. chatbot that would be able to answer employees' questions within the app. Companies like EY and Intel have been using chatbots for onboarding employees and answering HR inquiries for several years now, though they haven't expanded those platforms beyond internal use.

So far, Spoke has raised $28 million from VC firms including Accel Partners and Greylock Partners. Srinivasan says that to start, it's focusing on companies with 50 to 500 workers. He wouldn't reveal how the company plans to price Spoke.

He thinks that the company's cross-platform, easy-to-use approach will give it the edge it needs.

"We're trying to take a design-first, lightweight A.I. approach to a common problem that every company faces," he says. "I think the design and usability are as important as the machine learning itself. So that's our emphasis right now." ​

Published on: Feb 21, 2018