Persuasion Models Outperform Claude and ChatGPT | Dan Lee artwork

Persuasion Models Outperform Claude and ChatGPT | Dan Lee

MTS

November 8, 2026

Dan Lee, co-founder and CEO of Nooks, explains how their $500M non-dilutive capital facility from General Catalyst powers specialized AI agents to automate the revenue lifecycle.

0:00

Dan Lee

Because sales is more about persuasion, we're doing things like fine-tuning our own open-weight models on Nooks proprietary data that outperform Claude and ChatGPT on sales-specific benchmarks, on things like cost and quality and latency.

We're building actually RL environments for sales, because sales is a lot less verifiable. We're building basically these simulations where you can simulate what a customer, like how a customer might react. So you can actually test hundreds of scenarios to see which one's going to perform optimally.

0:35

SPEAKER_2

Hello everyone and welcome back to MTS. Today, I am joined by Dan Lee, the co-founder and CEO of Nooks, which just announced their $500 million in non-dilutive capital from General Catalyst to put AI agents next to every sales rep. So Dan, welcome to MTS.

0:52

Dan Lee

Thanks for having me.

0:53

SPEAKER_2

Thank you for joining. Let's talk a little bit about this $500 million in funding, because it's not typically what you see. What are you guys planning to use that money for?

1:02

Dan Lee

It takes a lot of capital to transform an industry, and sales is one of the oldest and largest professions in the world.

If you look around, everyone has seen that every knowledge job is changing. If you ask our engineers, we were just talking about this. Their job looks completely different than it did six months or a year ago, and the same has been happening in legal, in customer support, in other back office jobs. Sales has not changed as much, though, but it's actually a way bigger market. In the US., there are actually 12 million Americans who sell for a living, and it's more than all these other jobs combined.

It hasn't changed as much, we think it's because it's a harder problem. It's all about persuasion, and rather than reasoning, it's not as much a model capability. But the pace of the frontier, both from the labs, as well as our own applied research, makes us really confident that sales is going to be the next job to inflect.

1:54

SPEAKER_2

Yeah, I want to understand the larger landscape when it comes to sales, because my understanding of sales is, again, it's persuasion, but it's also your network, it's the people you have authentic relationships with. So how does AI come into play here?

2:08

Dan Lee

Yeah, I think in a lot of jobs, AI is making you want fewer people, but sales is interesting because when you can make a sales team more efficient, you actually normally want more sales, whereas in support, if you can make a support team more efficient, you normally don't want more support tickets. Or if you can make an accounting team more efficient, you normally don't want more audits.

And I give this example in recruiting. If you use AI to completely automate an interview process, and another company uses AI to actually save the interviewer's time, and which job would the candidate pick? They would probably pick the one where you have a better candidate experience. Sales is kind of like that, where you're building trust and relationships and actually solving real customer problems. So I think that part of the job is still human. That doesn't go away. But AI can do all the stuff behind the scenes to help people spend more of their time on those real human parts of the job.

3:03

SPEAKER_2

So where exactly does AI come into play here?

3:06

Dan Lee

Yeah, so I'll share it. Like Nooks, we are helping the best companies in the world grow faster with our revenue agent platform.

It helps sales teams, you know, it deploys sales teams, AI agents, across the revenue life cycle, from a customer coming inbound to the website, to like doing outbound sales, to growing and expanding accounts and closing deals. And our agents have access to all the same tools that a rep would, like writing emails and doing research, generating decks, even like answering technical questions on calls. And in a lot of cases, our agents are actually able to come up with better deal strategies than a rep would, because they have access to way more data.

3:46

SPEAKER_2

What do you mean more data? So how exactly does a human do it versus how an agent does it? And then what can the agent not replace?

3:53

Dan Lee

Yeah. So the agent, in most cases, is not replacing persuasion, right? I think there's like easy sales problems, like if someone comes to your website and mostly wants to buy your product, and you just need to answer a couple of questions to get them over the line, that's an easy problem that's going to be fully automated with agents. Where we have agents that can demo the product, that can talk and hold a full conversation with a customer, because that's an easy persuasion problem. But there are also hard persuasion problems where you have 20 stakeholders in a deal, and you haven't talked to 10 of them yet, they don't know what your product does, they don't have any budget, right? And that's a hard persuasion problem where that's going to need a human.

Read the other 11 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/1000800561990