AIVA From [24]7 -- AI That Finds Intent In Vague Terms

[24]7 introduced AIVA on Wednesday, an AI-powered Virtual Agent that integrates digital interactions with the web, messenger services or phone calls and understands inquires that are filled with vague terms.

For example, when a consumer calls a phone company, and rather than asking "why is my phone bill so high," says "I want to throw my phone across the room," AIVA can understand that the consumer is frustrated and may be looking for a new service plan.

Angela Sanfilippo, senior director of product marketing at [24]7, says 45% of consumer intent is vague or unrecognizable. Integrating AI can help the platform access data on the back end to help AIVA better understand the customers' intent and questions.

The virtual assistant works with both speech and digital media to eliminate silos between channels and to solve customer queries much faster. This allows companies to build the assistant once and integrate it into any self-service channel such as a brand's website, phone system or messaging platforms such as Apple Business Chat and Facebook Messenger.

The intelligent intent engine, which learns over time, is based on advanced data modeling from billions of prior customer interactions. It taps into [24]7's prediction-based modeling of 1.6 billion customer interactions and may recognize, for example, that a customer has already placed a call and is having a conversation with the chatbot or is on the website while they are placing a phone call.

Human behavior also comes into play. Generations display different behaviors, Sanfilippo says. Those age 35 and older typically reach for the phone, she said, citing data from the Dimension Data report, 2016 Global Contact Centre Benchmarking.

[24]7 is working with between 150 and 200 clients that use an informational VA, which means that consumers can ask a question and get an answer. Two clients are looking to implement AIVA, which integrates digital and Interactive Voice Response (IVR) for phone queries.

The platform also uses the same language intent model across speech and digital channels, so all interactions between the consumer and brand draw from the same business logic, application logic and back-end integration capabilities with enterprise systems such as CRM and billing.

As for the AI startups, they are finding it a bit more difficult to secure funding. CBInsights reports that the number of deals to AI startups in Q2 2017 fell 19% from the 297 deals seen in the quarter prior.

Despite the decline, second-quarter 2017 had the second-highest number of deals since the start of 2012. Top verticals for the quarter include healthcare and cybersecurity, while AI in fintech, sales and CRM as well as commerce saw slowdowns. 

Next story loading loading..