Co-founders: 3 (Stephen Campbell, Amin Akhtar, Xinchi Qi)
Employees: 4
Core Technology: AI-Powered Customer Segmentation for eCommerce; helping brands and agencies understand customer behavior by creating smarter segments for targeted email and paid acquisition campaigns.
Table of Contents
What's Your Story?
I grew up around a lot of entrepreneurial environments, so from a young age, I had a sense that I wanted to build something of my own. That really solidified when I started doing internships and working at big tech companies—like Ericsson, and I even had offers from Apple.
Those experiences made it clear that the traditional path wasn’t for me. It wasn’t just about being an employee—it was the structured, sheltered environments where everything is planned and things move slowly. That just didn’t suit me.
Revamp AI is actually my third startup. The first was in agri-tech focused on IoT sensors, and the second was Dreamwell, an AI-powered influencer marketing SaaS platform. Across those experiences, I probably made every early-stage mistake a founder could make. At Dreamwell, I had two co-founders and it was a valuable experience, but we struggled to find strong market validation early on. The traction just wasn’t there.
Over time, I personally discovered a more promising niche within marketing—one that had significantly more potential. At the same time, the vision between me and the team started to diverge. So I decided to exit and join a new venture—Revamp AI—focused on that clearer opportunity.
Biggest Struggle?
The biggest challenge for any early-stage startup is finding product-market fit. It’s easy to get distracted by customer noise—people suggesting features they think they want, but that often don’t lead anywhere.
Unless there's real commitment—ideally monetary—those requests can be misleading. At Dreamwell, we didn’t manage that well and got pulled in too many directions. It’s something we’ve significantly improved at Revamp AI, where we now stay laser-focused on solving validated problems.
What keeps you going?
What keeps me going is the belief that I can do this. But more importantly, you have to know what you're signing up for. Building a startup means going through a lot of pain—“eating shit,” as the saying goes—until one day you hit product-market fit and things start to change. If you don’t understand that upfront, it’s easy to burn out. You need to bet on yourself and truly believe in what you're building.
Early mistakes?
One of the biggest early mistakes is not clearly defining what kind of startup you’re building. You need to understand what product you’re creating, who it’s for, and how you plan to reach and sell to those users. That clarity should shape your entire strategy. Don’t stay stuck in your head, assuming what users want—talk to them, iterate quickly, and use that feedback to drive toward product-market fit.
If you're technical, that’s a huge advantage—but you also need to understand the mechanics of tech-driven product-market fit. There are great frameworks and books out there, but the key is applying them fast and early. That’s where everything starts.
The Problem: Poor, Impersonal Marketing Emails and Messages
Email marketers and paid media buyers often face major hurdles when trying to understand customer behavior:
eCommerce data is scattered across multiple platforms
Most brands and agencies lack the technical expertise to unify and make sense of that data
Even when they manage to consolidate it, turning the data into insights and action remains difficult
As a result, marketers rely on broad, generic segments for emails, while media buyers settle for basic targeting in acquisition campaigns—leading to millions of impersonal emails and wasted ad dollars on poorly defined audiences.
The Solution: Personalised Marketing Messages using AI
Revamp continuously tests message variations in real time and automatically delivers the top-performing content to each customer, maximising engagement without any manual effort. It enables true 1:1 behavioral targeting through dynamic wait-time optimisation, real-time message generation, and precise brand voice control—all fully automated. As customers interact with your messages, Revamp’s AI learns from their behavior and campaign results to refine and improve your copy over time, ensuring your messaging gets smarter and more effective with every send.
How did you come up with and validate the idea?
Revamp has pivoted a bunch of times. I actually joined in early 2024—around January. My co-founder found me through the Gen AI Collective, this community he started. We got to chatting about different problems we wanted to solve, and he was like, “Hey, you just moved to the Bay Area—why don’t you join us?” So I did.
When I came in, I looked at what Revamp had tried so far—things like an email segmentation tool, campaign builder, stuff like that. But it wasn’t really sticking. So I said, “Alright, let’s look at the users we already have, especially the ones not paying—what’s the one thing they actually want, and what’s something that would clearly show our value?” That led us to flow automations using AI. We started pitching that to users, and it just stuck. That was the first moment where it really felt like we had something.
I’d say at that time, our main KPI was just straight-up revenue, to be honest. So it wasn’t like we had a fixed number of users we planned to talk to or a super formal validation process. It was more like—what’s the idea that actually sticks and drives real growth and traction? That was the truth we were chasing. When I joined Revamp, there wasn’t much revenue, if any. So my number one goal was just to get the company alive—get something working that people would actually pay for. That’s what guided the decision.
What was your process for building the MVP?
That’s a pretty high-level question, so I’ll break it down into two parts.
