How Gridspace Applies Conversational Intelligence in Real Time
There’s a lot of useful information in conversations recorded by call centers. California-based startup Gridspace wants to help businesses explore and mine that data and insight to devise strategies to improve how the business functions in real time.
Conversational Intelligence
Gridspace’s flagship Sift platform at its core transcribes and examines conversations for real-time information about customer interactions. That data then becomes another tool to establish a picture of the company’s present and future. The artificial intelligence channels that insight back into the system, adjusting its responses to callers based on not only what they are saying, but the context of how they are saying it and in what circumstances.
“Short form command speech with a [smart speaker] is very different compared to natural conversations with customer service agents or friends or even how people would like to eventually talk with computers,” Gridspace co-founder and CEO Evan Macmillan told Voicebot in an interview. “Working on components for human-to-human speech led to Gridspace Sift. We can transcribe and annotate conversations, and ultimately automate routine tasks.”
Call center automation is a growing space for voice and AI technology. For instance, IPsoft’s Amelia platform and the virtual agents developed by Inference Solutions, as well as Google’s CallJoy service, are all designed to smooth the interaction between customers and call centers. Gridspace’s interface with customers, meanwhile, is born out of its focus on the layer above the point of interface with the consumer.
The value to organizations from aggregated conversational content can provide information about customers and trends. It becomes more than just a transaction; it provides intelligence to the business. Managers can see what’s happening in real-time and react more quickly to a broader set of information. That knowledge can then be filtered back into direct customer call center interactions through automation or used to create new features, products, and services.
Identifying Black Swans and Anticipating Needs
Macmillian co-founded Gridspace in 2012 with Anthony Scodary Nico Benitez. The co-founders worked with SRI Speech Labs, the research group behind the original IP used to create Nuance and Apple’s Siri voice assistant.
“We were really excited about technology that could make sense of speech,” Macmillan said. “We wanted to explore business insights in conversation.”
The company started with small corporate investments but has since raised $4.7 million from investors including Wells Fargo Accelerator, Bloomberg Beta, and USAA according to Crunchbase data and conversations with company executives. USAA is also a client, and one that offers a good example of a use case for Gridspace, Macmillan said.
USAA customers who were also government workers called the bank with questions and concerns about car payments, house loans, and related bills ahead of the government shutdown in 2017. With the data from the recorded conversations among other data sources, USAA decided to offer a zero-interest loan to customers affected by the shutdown, before it even began. Actionable information like this is what Gridspace’s customers are looking for, according to Macmillan.
“The megatrend really helping Gridspace Sift is the push to do more with data in the cloud. Even if [customers] aren’t moving all of their data into the cloud, they’re at least looking at data sets [as business assets],” Macmillan said. “They want to provide excellent customer service and get ahead of black swan events. We have to get that conversation into the cloud…Spoken conversation is something businesses must look at.”
Beyond Keywords
As the potential for augmenting business intelligence with conversational data becomes better known, there is more interest in what Gridspace can do, Macmillan said. Gridspace Sift’s ability to interact with different kinds of speech is especially useful in that regard. The software can identify and sort information based not only on keywords, but on how upset someone calling into a company’s earnings call might be, or even background noise like a crying baby that might increase the stress levels of a caller. All of that is grist for the analytics that can help steer a company’s fortune.
“Even folks who haven’t thought much about voicebots are realizing voice automation is not voicebots or nothing,” Macmillan said. “People are understanding that conversations shouldn’t be just turned into a file and put into a closet.”
You can learn more about Gridspace and its voice analysis solutions during an upcoming Voicebot webinar on August 22, 2019. Please register if you can join or if you would like a copy of the recording afterward. Register
Follow @voicebotai Follow @erichschwartz
