Podcast Episode

From No Idea to a $25M Raised in 3 Years | Mark Hughes co-founder Solidroad

The Entrepreneur Experiment

Episode 525Oct 1, 20261 hr 15 min

Who’s talking

2 speakers
Mark Hughes53 min · 73%
Gary Fox20 min · 27%

1Transcript601 turns · 1 hr 15 min

MH
Mark Hughes0:00

We believe that the future is hybrid, that it is human agents and AI agents.

GF
Gary Fox0:03

How does an AI-supported founder work in 2026?

MH
Mark Hughes0:08

It's not like we got in on our first attempt. Like, we got into YC on our fifth or sixth attempt.

GF
Gary Fox0:11

What's the biggest mistake you've made in the last three years?

MH
Mark Hughes0:14

I think not hiring people quickly enough. In an AI world where there's a lot of AI slop and Zoom fatigue, that in-person sales is gonna outperform and continue to outperform virtual sales.

GF
Gary Fox0:25

Mark, welcome to the pod.

MH
Mark Hughes0:27

Thank you very much for having me.

GF
Gary Fox0:28

You literally have flown in just for this. I feel special.

MH
Mark Hughes0:31

[laughs] Yeah, exactly. Just for this.

GF
Gary Fox0:32

It is just for this, right?

MH
Mark Hughes0:33

Yeah. I-- Also, we have a growing Dublin office, so I have to shout out the team in Dublin as well that I'm seeing a lot of people for the first time. So that office, last time I was here, there was three people. There's now 12, so lots of new faces.

GF
Gary Fox0:43

Oof. Okay. What do you do?

MH
Mark Hughes0:45

I'm the co-founder and CEO of Solid Road. We are an AI-first training and quality assurance platform for enterprise customer experience teams, which sounds like a lot of waffles. So to kind of give it the simplified version-

GF
Gary Fox0:58

You read my mind.

MH
Mark Hughes0:59

[laughs] When you call, uh, a customer support line or a call center, and they say, "This call is being recorded for training and quality purposes," historically, that was very laborious, where you'd have humans doing classroom-based learning with contact center agents and manually reviewing conversation transcripts with a scoring rubric to make sure people were following the right process. So we built AI agents to automate that.

GF
Gary Fox1:23

Okay. Makes a lot more sense.

MH
Mark Hughes1:25

Yeah.

GF
Gary Fox1:25

[laughs] Uh, okay. That's-- You've, yeah, you've kind of preempted my next question. That makes a lot of sense 'cause you often... Kinda one of those things you hear all the time but never really think why.

MH
Mark Hughes1:36

What happens with that?

GF
Gary Fox1:37

Is it only if someone, like, is like, "I'm not happy with the service. Can I get a transcript? I said this. I said I need to cancel my broadband six months ago"? I've heard of cases like that where they've pulled the transcript and go, "Oh, yeah, you did cancel it."

MH
Mark Hughes1:47

Mm-hmm.

GF
Gary Fox1:47

So that, that data is all used for training.

MH
Mark Hughes1:51

Yeah, exactly. So, uh, well, it should be. It's not. And, like, up until now-

GF
Gary Fox1:55

Mm

MH
Mark Hughes1:55

... for quality assurance in particular, um, companies would audit maybe 1% of total conversation volume, so-

GF
Gary Fox2:03

Okay

MH
Mark Hughes2:03

... that's all they're looking at. So yeah, there's lots of, uh, oversight that's missing from the other 99% of conversations.

GF
Gary Fox2:09

Okay. And give, give listeners a kind of s- context of where you're at as a company. So when did you start, and where are we now?

MH
Mark Hughes2:15

Yeah. So we founded the business in Dublin in early 2023. Uh, it was very much like myself and my co-founder, Mark and Patrick Inc. We have a memo about it, but kind of just wanted to work together. We didn't really have a clear idea what kind of business we want to build. We were both second-time founders. Um, kind of, uh, got the first product to market maybe later that year. Um, and then we started kind of having our kind of inflection point in, in 2024, and that's when we were, uh, accepted into Y Combinator. We did Y Combinator early 2025, and we raised our

2Episode notes

Mark Hughes started Solidroad in Dublin in 2023 with his co founder Patrick Finlay.

There was just one problem.

They did not actually know what they were going to build.

Three years later, Solidroad has gone through Y Combinator, built offices in Dublin and San Francisco, landed major enterprise customers and raised a $25 million Series A.

In this episode of The Entrepreneur Experiment, Mark breaks down exactly how it happened.

We go inside the early search for product market fit, how they got their first customers, founder led sales, getting into Y Combinator after multiple attempts, moving his life to San Francisco and the extraordinary speed at which Silicon Valley operates.

Mark also tells the story of meeting First Round Capital on a Saturday, meeting the partnership on Sunday and effectively having the investment agreed by Monday.

We then get into AI.

Mark explains how AI has completely changed the way he works as a CEO, why he believes repetitive work is most at risk, why customer support will remain a combination of humans and AI, and how he is effectively building an AI powered operating system around his company.

Finally, we talk about the biggest mistake he has made building Solidroad and the lesson he would give his younger founder self:

Move faster.

A fascinating conversation about ambition, fundraising, sales, AI, Silicon Valley and what it really

Showing a preview — retrieve the full transcript via the Exa API.

Powered by the Exa API

See something wrong or missing on this page?