Customer intelligence platform Enterpret Inc. today launched what it says is the first agentic feedback system designed to unify and act on customer signals across more than 50 data sources.

The service pulls in customer feedback from internal and external sources, including customer service platforms from Zendesk Inc., Intercom Inc. and Gong.io Ltd., and analyzes it in real time. It then organizes the data into a hierarchy of product and customer themes. Built-in “action agents” can alert teams to emerging issues or customer complaints as they happen.

Core features

The platform is based on two architectural elements. A Customer Knowledge Graph maps feedback to related entities such as user accounts, product features and business opportunities. “We can say 10 users give this feedback, and the sum of those opportunities amounts to $1 million,” said Varun Sharma, Enterpret’s co-founder and chief executive.

An Adaptive Taxonomy continuously adapts to evolving user language and product structure. “We’re automatically able to understand the hierarchy of your product,” Sharma said. “Level one may be a Zoom chat and inside of that are messaging, collaboration tools and administration. As new feedback comes in, it is automatically slotted in to the appropriate level and refreshes on its own.” Enterpret said this eliminates the need for teams to tag or retrofit templates manually.

Agentic capabilities

The new features are designed to move beyond analytics to automation. A natural-language interface allows users to ask questions like “Why are signups stalling in Europe?” and receive contextual responses sourced from live customer data.

The system also identifies potential problems, such as bugs or account-level churn risks, and routes alerts accordingly. Sharma described the approach as agentic because of its ability to ingest, structure and act on feedback independently.

“We’re automatically able to structure the feedback into a five-level hierarchy,” Sharma said. “You can ask questions in natural language, get answers very easily, or we have action agents which can detect bugs in real time.”

Enterpret doesn’t track implicit behavioral data such as mouse clicks or web navigation paths, focusing instead on written and verbal communication. Support for audio and video inputs and automated transcription is included.

No extra cost — for now

Sharma said the new capabilities are included in the existing product and are available to all customers on an annual subscription model.

“We’re trying to maximize to value we can deliver to as many customers as possible,” Sharma said. “But it’s a fast-evolving field so we may have to play around with monetization.”

The company said its service complies with System and Organization Controls Type 2 security standards and scrubs personally identifiable information at the edge unless requested otherwise.

Founded in 2020, Enterpret has raised $25 million in funding from Kleiner Perkins Ltd., Canaan Partners LLC, Wing Venture Management LLC, Peak XV Partners Operations LLC and Recall Capital Fund LP. Sharma said it is operating at a “seven-figure annual recurring revenue” rate.

Nick Bell, CEO of Enterpret customer Fanatics Collectibles AD Inc., said the software helps his team move faster. “Feedback doesn’t arrive in tidy survey results; it pours in from dozens of channels simultaneously,” he said in a statement. “Enterpret transforms what was once overwhelming noise into clear, real-time insights we can use to drive action.”

Image: Enterpret