telli Raises $15M to Build AI for Customer-Facing Operations

A customer call rarely ends when the conversation does.

Someone still has to update the CRM, schedule the appointment, send the confirmation, assign the request and make sure the next action happens. For many consumer businesses, that operational work remains spread across call-center software, ticket queues, dashboards and manual follow-ups.

Berlin-based telli has raised $15 million in Seed funding to bring those disconnected tasks into one AI-driven customer operations platform. The financing lifts the company’s total funding to $18.5 million.

The round was led by redalpine and included Mutschler, strategic investors and several angel backers. Existing investors Cherry Ventures and Y Combinator also participated.

Founded by Sebastianos Hapte-Selassie, Finn zur Mühlen and Philipp Baumanns, telli operates from Berlin and San Francisco. The founders previously gained direct experience with high-volume customer operations while helping scale Enpal’s heat-pump business.

The call is becoming an executable workflow

telli began with AI phone agents, but its ambition now extends beyond automating speech.

Its platform allows B2C companies to deploy agents that answer calls, qualify leads, schedule appointments, run follow-ups and resolve customer requests. The system is also designed to work across chat, SMS, WhatsApp and email as consumers move between communication channels.

The important distinction is what happens after the agent understands the customer.

A lead should not simply receive an answer; it may need to be scored, entered into a CRM and routed to a salesperson. A service request may require calendar access, a confirmation message and an update to another internal system.

telli is attempting to turn each conversation into a completed business process rather than another transcript waiting for human action.

Charlie manages the agents behind the agents

At the center of the platform is Charlie, an AI coworker for customer operations teams.

Charlie helps users configure agents, connect systems, review conversations, identify recurring problems and test changes to improve performance. The goal is to make AI agents easier to operate after deployment, when real customers begin introducing unexpected questions, interruptions and edge cases.

The company says its agents have already processed millions of conversations for businesses ranging from smaller companies to large enterprises. Named customers include Sky, Enpal and Vaillant. These adoption figures are based on company information.

Consumer companies are losing control of the channel

Customer communication used to move through a relatively predictable set of touchpoints. That structure is becoming harder to manage.

Some consumers call. Others use WhatsApp, email, forms or chat. Personal AI assistants may soon contact businesses and negotiate services on the customer’s behalf.

Supporting each channel through a separate tool creates fragmented histories and repeated work. telli is betting that companies will need one AI operating layer capable of managing the conversation regardless of where it begins.

That makes voice the entry point, not the final product.

The hard conversations will decide whether the model works

The new capital will support hiring across engineering, go-to-market and customer operations, with more than 15 open roles announced alongside the round.

telli’s next challenge is not producing another natural-sounding demonstration. It is handling the calls that do not follow a clean script: unclear requests, frustrated customers, interrupted conversations and tasks that cross several internal systems.

Businesses will evaluate conversion, resolution rates and operating cost. Customers will use a simpler standard: whether the AI helps them finish what they contacted the company to do.

Source : telli Official Announcement