ElevenLabs' AI Sales Agent Handled 3,000 Leads in 32 Languages

ElevenLabs' inbound sales agent has processed more than 3,000 customer leads this year, qualifying over 90% of them in 32 languages with no human intervention. Alex Holt, the voice AI startup's chief technology officer, shared the figures as evidence that agentic AI already delivers measurable value inside the company.

The sales development representative agent is one of several deployed across ElevenLabs. A coaching agent role-plays customer interactions with new sales hires, shortening onboarding. A recruiting agent conducts first-round interviews with job candidates. Inside the company's Slack, agents answer employee questions about internal processes on demand.

"I think what's most exciting about this space is that you know you always feel like you're on the tip of the iceberg," Holt said. He added that today's efficiency gains "in six months time, you can see even more gains."

How big is the agentic AI shift?

Agentic AI's share of total AI value grew from 17% in 2025 to 22% this year, according to research by consulting firm BCG. The report projects that share will reach 39% by 2030. Most surveys, however, find businesses remain ill-prepared to redesign workflows and orchestrate multi-agent processes, where one agent autonomously hands work to another.

Holt says most ElevenLabs employees work alongside agents, and a majority have built their own agents to absorb repetitive tasks. The business objectives stay fixed. Sales teams must use AI to generate more revenue faster. Customer service teams track satisfaction scores.

"It's exactly the same business metrics," Holt said. "This has to be an initiative that the business units themselves are driving and pushing. The agents and tech are a tool that they can leverage to do that."

What is the organizational setup?

ElevenLabs has placed a forward-deployed engineer on every team, and every group stays under 10 people. In recruiting, two engineers work in a pod with HR colleagues. Their single objective: cut the time it takes to move candidates through interviews.

Almost all applicants now get an initial screen from the AI recruiter. "That's 30 minutes that the recruiter can spend at the end of the process, convincing you to join," Holt said.

Holt joined ElevenLabs in September 2023 as an engineer, after five years at Palantir Technologies. He became CTO in June 2026. Three months later, the startup announced it had doubled its valuation to $22 billion. CEO Mati Staniszewski recently indicated the company may start preparing for an initial public offering in 2028.

Can agents hand work to other agents?

The next frontier, Holt says, is multi-agent systems—more shared context, more complex handoffs between agents. That work is already live in sales. The SDR agent uses voice AI to gauge a prospective buyer's seriousness, then loops in a human representative after screening.

Recently, one ElevenLabs customer closed a sale entirely through two agents. The SDR agent fielded the inbound request. An account executive agent then handled the legal documents and contracts.

"The value of that is you can now automate that whole chain for customers where it's the simple, most repetitive mechanical cases," Holt said. Employees, he added, can spend more time on complex requests and upselling.

What happens when agents fail?

Agentic AI demands patience. Holt asks whether the recruiting agent hires the right people and whether service scores improve with AI. "Some of these metrics can take a long time," he said. "We lean on the business owners to identify those leading indicators and the metrics we should be measuring."

Failures also occur. The industry has watched recent high-profile cases where OpenAI and Anthropic agents acted in unintended ways. ElevenLabs built a human-led monitoring and observation layer to govern its agents, but mistakes still slip through. One customer service agent issued a refund under $30 without escalating to an employee. The company reproduced the call, traced the agent's exact chain of reasoning, and updated the prompts and guardrails.

That audit trail is a hidden benefit, Holt argues. A human making the same mistake would likely have gone unnoticed.

"It's not just automating processes; it allows you to collect more data," he said of agentic AI. "I expect over the next few years, the trend that we're going to start seeing is people talking much more about optimizing these processes."