How ElevenLabs Turns AI into Measurable Business Outcomes

**Alex Holt** (0:00)
I can guarantee if you sign up right now, by the end of the afternoon, you'll have an agent that is doing something impressive and without having to spend, invest lots of time.

**Peter High** (0:11)
Welcome to Technovation. I'm Peter High, president of Metis Strategy. My guest today is Alex Holt. Alex is the first ever field chief technology officer of ElevenLabs, the AI voice and communications platform whose technology powers customer-facing agents for some of the world's largest enterprises. The company is four years old and has already achieved a valuation appropriately enough of $11 billion. In his role as field CTO, Alex leads the company's forward deployed engineering function, embedding teams directly with clients to bring voice AI deployments to life at scale. He joined ElevenLabs in 2023 as employee number 11, bringing with him deep experience building and scaling enterprise go-to-market teams at Palantir.
I look forward to hearing more about his journey and more about the future of ElevenLabs through this conversation. Alex, welcome to Technovation. It's great to speak with you today.

**Alex Holt** (1:02)
Thanks for having me. It's great to speak to you too.

**Peter High** (1:04)
Wonderful. Well, take it from there if you would about ElevenLabs. I'd love to have you describe it in your own words, a fascinating company that has grown remarkably over the course of just a little bit more than four years. Talk about the business you're in, if you would, Alex.

**Alex Holt** (1:17)
Yeah, of course. So ElevenLabs is a research and product company.
And our mission is to help enterprises to communicate better with their employees, with their customers, with the world.
And so really where we started is, very much as an AI research lab, building models to help technology to communicate. So the hardest problem and the one that we saw first was creating models that sounded, re-created the human voice. And this was the very first models, of a way you could hear them and they would sound indistinguishable from humans, which was a really, really impressive technical breakthrough. And on top of that, we then built increasing levels of product to make it easy to deploy. Those models are new subsequent models that we've built against real-world use cases so that enterprises can actually solve their communication problems. So today we have three main products. We have the 11 agents product, which enables customers to build voice agents from zero to complete very, very quickly. So these are for building virtual sales assistance, virtual customer support systems, virtual learning and development tutors, et cetera. So think agents you can call, you can talk to, you can chat to using text that are connected to all of your enterprise source systems and knowledge bases and can have real effects like refunding customers or booking appointments. And they sound and feel human in the same way that the most empathetic customer support or sales agent would do.
And so that's kind of our first product line. The second product line is 11Creative, which takes our model technology and combines them with tools that make it very easy to produce content. So you want to create podcasts in 100 different languages, or you want to create, take video content or podcasts you would have created and dub those into other languages, but use the original actors' voices, but suddenly have them speaking Polish or Spanish. Or if you want to create entirely new content, so creating personalized ads for every single one of your listeners, depending on their demographic, using our AI models and other AI models that exist, and then you can do that in those tools. So that's solving the communication problem of how do we communicate with my broader audience and the world through those pieces of content. And then the other thing is that we still make all of these models available to developers through 11API, which is our third pillar, which is we are, at our heart, one of our big differentiators is we have AI researchers that are building these models that are groundbreaking, and we want to ensure that people can access those as developers as well.

**Peter High** (4:09)
And of the use cases that you've described across 11Agents, 11Creative, 11API, as you think about the use cases, how much of it is specific to an individual, like me speaking to my team, my customers, the world, to use your earlier point of point, versus any voice you use. So just a human voice that is very realistic to speak to a variety of customers who are calling in about different inquiries of a company, for example.

**Alex Holt** (4:37)
Yeah. So it's very configurable. A large part of one of the things I'll talk about is that we believe that every enterprise and every use case will want a slightly different application of this technology. And so what our technology allows you to do and what the models allow you to do is they can you can create entirely synthetic voices. You can create voices of clones of your own voice or voices that you have the rights to do so.