Customer engagement startup Pathlight today rebranded as Echo AI and introduced new features for its flagship software platform.

The company, which is incorporated as TTSF Inc., is backed by $35 million in funding from Kleiner Perkins and other investors. It provides a platform that uses artificial intelligence to find areas for improvement in call center teams’ interactions with customers. The platform can, for example, identify cases where a user’s technical support question wasn’t fully addressed.

Echo AI doubles as a marketing tool. According to the company, its platform extracts information on customers’ buying preferences from contact center calls and makes the data available for use in promotional campaigns. The new capabilities Echo AI introduced today in conjunction with its rebrand enhance this part of its platform’s feature set.

The first addition is called Conversation Actions. According to the company, it detects cases when a customer expresses interest in buying a certain product during a call. The feature can then send that customer a personalized sales offer through a third-party marketing automation service.

Echo AI provides integrations with marketing automation tools from Salesforce Inc., HubSpot Inc. and several other companies. It can likewise send information about shoppers’ buying preferences to so-called customer data platforms, or CDPs. Those are applications that marketers use to aggregate data about consumer behavior from different systems and find ways to improve ad campaigns.

Besides sending upsell offers, Echo AI’s new Conversation Actions feature can also automate several other tasks. It’s capable of notifying marketers when a certain metric, such as the number of users who mention a competitor during calls, exceeds a certain threshold. Additionally, Conversation Actions can offer incentives such as discounts when a user is considering canceling a subscription.

As part of today’s update, Echo AI is also expanding the range of data sources that its platform draws upon to understand shopper behavior. Until now, the platform focused mainly on processing contact center calls, emails and texts. Echo AI can now also ingest app reviews, social media posts, net promoter score surveys and chatbot conversations.

“The features we’re announcing today allow us to elevate Conversation Intelligence beyond the contact center, so that every department can benefit from our insights and actions,” Echo AI Chief Executive Alex Kvamme wrote in a blog post today. “We believe that every employee should be able to access and benefit from conversation intelligence, as every employee plays a role in serving customers.”

The platform visualizes the customer behavior patterns it finds in graphs to ease analysis for workers. Echo AI can, for example, generate a pie chart that highlights the most common reasons users cancel a subscription. The platform also tracks buyer sentiment and a range of related metrics.

Under the hood, Echo AI uses large language models to analyze the customer data it ingests. It leverages both open-source LLMs and commercial models from providers such as OpenAI. Customers can optionally fine-tune the LLMs by training them on internal marketing datasets. 

“We’ve built infrastructure that is capable of analyzing millions of conversations, across dozens of dimensions, using many LLMs at any time, and with unparalleled accuracy and consistency,” Kvamme detailed. 

Image: Echo AI