Twilio Inc. said today it’s infusing its customer engagement tools with a hefty dose of artificial intelligence.

At its annual user conference SIGNAL 2023, it announced new predictive and generative AI capabilities to drive deeper insights into customers and power recommendations and suggestions for developers.

Twilio made its name for its application programming interfaces, which make it simple for developers to plug voice, test messages and video capabilities into their applications and streamline communications. In addition, it sells tools for customer engagement such as its Twilio Engage growth automation platform, which helps marketers improve customer relationships with personalized experiences.

As the name suggests, Twilio’s new CustomerAI capabilities are mostly focused on the customer engagement side of its platform. They’re designed to help companies expand their perception of customer data and activate it more easily, the company explained.

Predictions is a new predictive AI capability that’s designed to help marketers create “hyper-targeted audiences,” personalize multichannel experiences and trigger superior customer journeys. It analyzes each customer to make predictions about their lifetime value and their likelihood of repeat purchases or churn, based on data within Twilio Segment. Twilio said Box Inc. is already using Predictions to optimize the performance of its marketing campaigns and unlock additional revenue opportunities.

Meanwhile, Voice Intelligence is a new capability for Twilio Flex and Twilio Voice that can be applied to customer service channels. It relies on natural language recognition to listen into customer calls and report common feedback, competitive insights and compliance risks.

It can also help to manage regulatory requirements with automated personal information redaction. According to Twilio, one organization that tested Voice Intelligence in beta said it was able to improve lead attribution management margins by 21%.

Generative AI

On the generative AI side, Twilio announced new functions that are meant to help activate customer data and power adaptive engagement in Twilio’s Engage, Flex and Segment tools.

Twilio said they include a new Generative Email Design tool for Twilio Engage, which will allow marketers to enter simple text prompts and transform ideas into HyperText Markup Language or HTML instantly. The company explained that this builds on existing AI tools in Engage such as Smart Headlines, Smart Buttons and the Smart Image Generator, which suggest conversion-worthy content marketers can add to their email communications.

As for Generative Journeys, this helps to eliminate the manual processes when architecting customer journeys, Twilio said. With it, marketers can just describe the type of campaign, be it promotional or win-back or something else, define their audience and the channels they wish to use, and Twilio Engage will automatically create the customer journey.

Twilio Flex is getting a new GenAI Agent Assist feature that will create recommendations for contact center agents. It will suggest “next best actions” during each interaction with customers to try and improve outcomes. At the same time, it will also summarize each customer interaction to help the agent wrap things up in the best possible way, Twilio said.

In addition, Twilio Flex’s Flex Unify tool that’s currently in beta will soon launch in general availability to natively surface customer profile data from Segment in real time. This will help agents to interact with customers in a more personalized way.

Twilio Chief Executive Jeff Lawson said that AI delivers more value when it’s paired with first-party data sets. “This is where Twilio is most differentiated,” he explained. “With Customer AI, artificial intelligence becomes customer-aware.”

AI-ready data

Alongside the new AI tools, Twilio announced what it said is a “Composable, AI-Ready Customer Data Platform” in the shape of Segment B2B Edition, which is based on a unique zero-copy architecture.

Twilio explained that to get the best out of AI, companies require a clean and consistent data foundation. Segment B2B Edition is meant to provide this, incorporating LinkedIn profiles to help businesses build up a graph of customer relationships, linking their events, experiences and profiles with the additional data held in Segment.

It provides a better line of sight across data that’s graphed to each customer, and makes it easier for companies to train more accurate AI models and build advanced segmentation, Twilio said. They can then create more personalized recommendations and experiences.

Andy Thurai, vice president and principal analyst at Constellation Research Inc., said the graph connection for LinkedIn profiles is perhaps the most compelling and interesting feature announced today. “It can potentially unearth certain events and make meaningful personalized marketing messages or offers that might be timely and appealing,” he explained.

As for the zero-copy architecture, it’s designed in partnership with Databricks Inc. to ensure their data is AI ready. Twilio said it’s integrating capabilities from Databricks to enable users to query data warehouses or data lakes directly, removing the need to copy data and perform extract, transform and load. By doing this, it eliminates the problem of redundant data and generates insights more quickly, Twilio said, while reducing the need to build and maintain data pipelines.

Thurai said the problem Twilio faces in trying to combine large language models with proprietary customer information is that that it’s own customer data platform is not always up to date. Therefore, Twilio is aiming to query third-party data lakes in real-time to get additional information that enriches the data from its own CDP.

“This will eliminate the need to copy data and perform time-consuming ETLs,” Thurai said. “It’s a tall order to accomplish, but they have a good partner in Databricks, so it may work out.”

Still, the analyst said, Twilio will find it tough to move the needle in an area where it faces lots of tough competition from rivals such as Salesforce Inc. “It’s a good step in the right direction, but many of Twilio’s competitors are doing the same or even better things in this area already,” Thurai pointed out. “It’ll be interesting to see what traction it gains with these initiatives.”

Photo: Twilio