Cloud-based communications platform provider Twilio Inc. today announced the launch of a new set of tools for developers designed to help them communicate more reliably with customer across multiple channels using artificial intelligence and machine learning technology.

In order to address the needs of developers to create apps and portals that allow natural, personal communication with customers, Twilio has announced two new application programming interfaces that will help ease the creation of such communication channels: Conversations, a unified API that allows developers to create cross-channel group communication at scale, and Media Streams, an API that gives developers tools to analyze voice calls in real time in order to apply intelligence to enable better customer outcomes.

“Over the last two decades, we’ve watched businesses evolve their communications with customers from the phone call, to website chat, to native mobile apps,” said Chee Chew, chief product officer at Twilio. “Leading companies have figured out that the next evolution of great customer experience is through messaging. Twilio Conversations empowers businesses to build personal, long-lived connections with their customers on the channels they prefer.”

According to Twilio, 9 out of 10 customers globally want to be able to converse with brands, but the current tools to enable brands to hold conversations with customers can be extremely complicated because there are lots of ways to interact. Channels such as voice, SMS, chat apps and web pages all add increasing complexity.

To address that complexity, the Conversations API and Twilio’s infrastructure enable developers to create conversations with users that can cross separate channels and maintain state with cross-channel group communication. Channels that the Conversations API can manage at scale include short message service and multimedia message service (text messaging protocols), general chat and WhatsApp.

Using Twilio Conversations, developers would be able to build secure, compliant conversations for many enterprise concerns including financial services. For example, a customer might want to converse with their personal wealth advisor on text through a company platform but also want to run the same idea by a partner who is overseas and on WhatsApp. With the Conversations API, it would be possible for Twilio to enable a conversation between the banker and you on text (still secure and complaint) while also including your partner on WhatsApp in the conversation.

Twilio Media Streams will tap into the Twilio Voice platform with an API that access the raw audio stream of any Twilio-powered phone call for secure analysis of the conversation in real time.

According to the company, a majority of conversations with any business happen through contact centers and happen in voice. Most of the data generated from voice conversations is only reviewed during post-call recordings or manual human interaction (basically a manager listening in).

With Media Streams, a powerful machine learning algorithm provided by a Twilio partner can “listen in” and help deliver additional assistance to customer service employees at a contact center during the call itself. To do that, Twilio partnered with Amazon Inc., Google LLC, and GridSpace Inc. to get access to advanced machine learning capabilities.

With Amazon Lex, developers can have a machine learning algorithm “listen in” and help determine the intent of a caller. The platform uses a conversational AI to understand human speech and provide suggestions based on context. It’s the same platform that powers the virtual assistant Alexa.

Attaching Google Text-to-Speech through the API can automatically transcribe the conversation in real time as well as make suggestions to customer service agents based on the contents of the call and the company’s built-up knowledge base.

Finally, Gridspace’s insight analytics can be used to adjust based on the content and tone of the caller in real time by automating the creation of labels, scores and context analysis.

“Media Streams ensures our customers have great experiences when they call into Globo’s telephone interpreting services,” said Jonathan De Jong, vice president of engineering at Globo, a multilingual automatic language translation service. “By using real-time sentiment analysis to monitor each customer experience in real time, Globo has dramatically increased call center productivity across our thousands of agents and interpreters.”

Both the Conversations API and Media Streams API is available in public beta today for Twilio customers.

Photo: Pixabay