The numbers tell one story. A $15 million seed round, announced on July 23, 2026. Customers like Sky and Viessmann. Lofty claims about efficiency gains and cost reductions. But for Finn zur Mühlen, the pitch for telli begins somewhere more mundane: a year spent scaling a heat pump call center from essentially nothing to 150 agents.
That operational slog—managing appointment flows, chasing down leads, coordinating service calls across Germany—left zur Mühlen and his co-founder Philipp Baumanns convinced that phone-based customer service was ripe for an overhaul. Not incremental improvement. A ground-up rebuild using voice AI.
"You live through that chaos, and you start asking why so much of it needs humans at all," zur Mühlen said in a recent interview, though he's quick to clarify that telli isn't about eliminating jobs wholesale. It's about automating the repetitive, high-volume interactions that burn out agents and drain budgets.
The Berlin-based startup now has $18.5 million in total funding—a $3.6 million pre-seed announced in April 2025 and this latest $15 million seed round led by Swiss venture firm redalpine. Cherry Ventures, Y Combinator (telli graduated from the accelerator's Fall 2024 cohort), and Mutschler Ventures all participated, alongside a handful of notable strategic angels: Marc Bitzer, CEO of Whirlpool and a BMW supervisory board member; Alexa von Bismarck from payments giant Adyen; and Martin Koehler, formerly a director at Flixbus and Lufthansa.
Fifteen months separated telli's pre-seed from this seed close. In the current fundraising climate, that's brisk.
The Product Thesis
At its core, telli builds AI voice agents designed to replace—or at least augment—traditional contact center platforms like Talkdesk and Aircall. The system handles both inbound and outbound calls: appointment scheduling, lead qualification, payment reminders, collections, and first-line support triage.
The company offers what it calls telli Studio, a no-code builder that lets customers configure agents without engineering help. Voice cloning technology allows brands to maintain consistent audio identities. Support spans more than 20 languages. Native integrations with Salesforce and HubSpot smooth the handoff to CRM workflows.
One feature stands out, if mostly because it's named: an AI coworker called "Charlie" that reviews call transcripts and iteratively refines agent behavior. Think of it as an automated quality assurance layer, continuously adjusting scripts and response patterns based on what's working—or not.
For regulated verticals like financial services, telli has built compliance scaffolding around call recording, disclosures, and consent management. Whether that's robust enough to satisfy legal teams in heavily scrutinized industries remains to be seen, but the startup is clearly signaling ambitions beyond straightforward use cases.
The Team Behind It
Beyond zur Mühlen and Baumanns, the third founder is Seb Hapte-Selassie, who studied computer science at Stanford and previously held engineering roles at Pitch, N26, and BCG Digital Ventures. That's a fairly typical European startup founding trio: operators who've seen the pain firsthand, paired with a technical co-founder who can actually build the solution.
The company's employee count hovers around 20, though that figure may have shifted since it was last reported. What's clearer is the revenue trajectory: telli claims 9x year-over-year growth. Of course, "9x" means little without knowing the starting point. A leap from €100,000 to €900,000 tells a different story than €1 million to €9 million. The startup declined to share specifics.
Customer Traction and Case Studies
telli lists Enpal, Sky, Viessmann, Vaillant, Bark, and Homeday as customers. Some of those relationships come with quantified outcomes attached—though, as always, vendor-supplied metrics should be read with a degree of caution.
Enpal, where zur Mühlen and Baumanns previously worked, reported a 30% jump in weekly service appointments and a 40% reduction in time spent on dispatch logistics. Bark, a UK-based lead generation platform, added 800 hours of weekly calling capacity and cut cost per contact by 70%. Those are the kinds of numbers that get investors' attention, even if the full context—baseline volumes, implementation timelines, margin impact—remains opaque.
According to zur Mühlen, the seed round came together relatively quickly, propelled by customer referrals and rapid expansion within existing accounts. That kind of word-of-mouth growth is harder to manufacture than top-down enterprise sales, which gives it a certain credibility.
A Crowded, Capital-Hungry Space
Voice AI agents have become one of the venture community's preferred themes. Simple AI secured a $14 million seed in February, led by First Harmonic. Vapi closed a $50 million Series B in May, pushing its cumulative funding to $72 million. There's clearly an appetite—among both investors and businesses—for products that promise to automate voice-based interactions.
Market research firms have responded predictably. Grand View Research projects that the voice AI agent market will balloon from $2.54 billion in 2025 to $35.24 billion by 2033. Whether that forecast proves prescient or wildly optimistic is anyone's guess. These projections often assume smooth adoption curves and regulatory environments that remain cooperative. Reality, as usual, tends to be lumpier.
telli didn't disclose its valuation, which is fairly standard for seed-stage companies. The startup operates out of Berlin and San Francisco, targeting sectors where phone-heavy operations are table stakes: solar and HVAC, real estate, recruiting, financial services, healthcare, telecom.
What's Next
For now, telli is focused on expanding its customer base and refining its compliance capabilities. The company is betting that as voice AI models improve—and as businesses become more comfortable delegating customer interactions to software—the total addressable market will expand beyond early adopters.
Whether that bet pays off depends on factors the founders can't entirely control. Regulatory scrutiny around AI-generated calls is intensifying in several jurisdictions. Consumer tolerance for automated interactions varies widely. And competitors are circling, many with deeper war chests and existing enterprise relationships.
But zur Mühlen and Baumanns have at least one advantage: they've been on the other side of the phone line. They know what it's like to manage a sprawling call center operation, to juggle KPIs and agent churn, to wonder if there's a better way. Now they're building it.
