How You Deliver Words Matters More Than What AI Says | Alex Holt artwork

How You Deliver Words Matters More Than What AI Says | Alex Holt

MTS

November 12, 2026

ElevenLabs Field CTO Alex Holt breaks down their major Eleven Eleven Summit launches, including the new Interaction Engine, agent orchestration, and the NVIDIA Nemotron partnership.

0:00

SPEAKER_1

Hello, everyone.

0:00

SPEAKER_2

Welcome back to MTS. Today, we are joined by Alex Holt, the Field CTO at ElevenLabs, and one of the company's first employees. Today is its summit, ElevenLabs Summit at Eleven Eleven. The day, awesome day. I know you guys just launched Interaction Engine, new capabilities for Eleven Agents platform, and a healthcare transcription model. Alex, welcome to MTS.

0:27

Alex Holt

Thanks for having me. It's great to be here.

0:29

SPEAKER_2

Great to have you as well. Lots of big announcements that you guys had today at the summit. For people who are not there, who are just tuning in, what did you guys announce?

0:39

Alex Holt

We announced a bunch of stuff here at Eleven Eleven. Really what most of our announcements are focusing on is thinking about how agents and humans interact, and building out that interaction layer for AI agents. And so we announced a host of new models and the ability to fine-tune models to be more performant in your domain, in your industry. We announced tools to make agents remember what's happening between conversations, agent memories. We announced the ability to take work outside of those conversations to do it before and after conversations, through something we're calling routines.

We announced the addition of Nemotron models coming onto our platform.

We're deepening our partnership with Nvidia. We also announced and gave a sneak preview of something calling Eleven AI, which enables the interaction, not just between humans and AI, but supercharging interactions between humans and other humans using AI, which is also a very new space for us.

1:43

SPEAKER_2

Man, this is so fascinating. Lots to unpack. One of the things I wanted to start off talking about is you say, you're helping how humans interact with agents. Agents, plural, because right now, many humans, especially at the enterprise layer, are interacting with many different agents, and you guys are working on some tools for orchestration at that layer. Tell me a little bit more about that.

2:05

Alex Holt

Yeah.

When you're interacting with AI, with agents, we fundamentally believe that they shouldn't be just an isolated instance. Actually, if you think about what you're trying to achieve when you're an enterprise is having an interaction with a customer, there's going to be multiple stages to that, there's going to be multiple steps. And so really what we're doing is we're building the tool set and the platform so that you can empower that entire user journey and not just a single point of contact. And so that may mean you have, if you think about a sales process, you have a qualification process which may require one set of skills and an agent to run that qualification. If that qualification is successful, then you're going to want to then stop the sales process proper in which you send over to an account executive. And you may want to do that with another AI agent.

It's really important you don't lose context between those two agents and those two agents can call on the context they've had in the previous conversations about the customer so it feels seamless.

Depending on how that final conversation with an AI account executive goes, they may then want to hand over to a human to finish the deal and take it to the next step, in which case they need that context and they need to understand the history of your journey with that prospect so far. Or maybe you actually then want to go and take action. Maybe you'd manage to close the deal using an AI agent and then you need to go and send a contract via a routine and do work outside the agent as well. And so these are the kinds of problems that we're thinking about. It's not just a single interaction and an AISDR, but it's actually thinking about what a sales process looked like in an AI native organization. It's multiple agents, it's humans being a core and reintegral part of that workflow as well. And we think that the next unlock is how do you optimize that entire process. Multiple agents, humans, processes that are happening outside of the agents. And you need to do that in a seamless way so that the end customer, the prospect, they don't know any of that complexity at all. It's just like a very well-oiled machine that just works.

4:17

SPEAKER_2

Yeah, no, it's really fascinating because as you mentioned, AI is highly intelligent. And yet, when you're trying to get it into the enterprise, when you're trying to get AI to just communicate with people, work alongside people, there's a huge gap. So tell me a little bit about your guys' plans on how you plan to close that gap.

Read the other 13 minutes of this episode here.

$15 gets you this episode and 99 more. Credits never expire. After checkout, one link brings you back here to read it.

Not what you expected? Full refund within 14 days.

Thousands of transcripts fetched by people building searchable podcast archives

Fetch transcripts with code

The same credits work from a script or an AI agent. The demo key returns a sample episode in full, no card needed:

request
curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.

Using your own key:

request
curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/1000801371448