Instead of layering automation onto outdated workflows, Dexter applies AI across the procurement lifecycle, from supplier evaluation and request intake to supplier management, invoice validation, and spend analytics, while addressing Europe-specific needs like regulatory complexity, data residency, and supply-chain transparency.

Based in Berlin, the company serves manufacturing, wholesale, and industrial businesses that are running into limitations with enterprise solutions. By integrating and adding their software on top, Dexter is offering their customers ways to extract more insights and is focused on deepening its product in core European markets with an emphasis on clarity, usability, and better decision-making.

We caught up with Bella Wu (CEO) and Caterina Kiehntopf (COO) to find out more. Join us as we talk about their journey from Dance, to Y Combinator to rethinking procurement.

About Dexter

Dexter is an AI-first procure-to-pay platform built for European mid-sized companies, founded by former operators – Bella Wu, Caterina Kiehntopf and Vincent Garrigues-Geithner who set out to fix the limitations of legacy procurement by helping teams understand what they buy, who they buy from, and how those choices compound over time. 

A new approach to procurement

Right now, for European procurement teams, procurement can be a lengthy and complicated process. Invoices require manual review. Data is unstructured and hidden in invoices, orders, contracts, and emails. And it can be difficult to extract meaningful information from enterprise solutions. 

Dexter was built to use that unstructured data and make the entire procurement process easier.

The company did not begin as a broad platform vision. It started with a very specific operational problem: receiving invoices, validating them against contracts, POs and Goods Receipts, and making sure the underlying processes actually worked. Once that problem was solved, something larger came into focus.

Bella explained that their first customer simply needed help handling invoices properly. They originally started, she says, with a customer struggling to receive invoices, check that they were being charged correctly, and make sure processes were in place — and that became the first product they built.

As the team improved at extracting line-item data, something larger began to emerge. Once they became good at parsing each line item and auto categorising these, Bella recalls, they realized how much insight was sitting inside them for procurement executives. Suddenly, she says, you could see what a company was really spending money on.

Invoices became a lens. Not just a record of past transactions, but a real time view into what companies were actually paying, line by line, rate by rate. Instead of relying on purchase orders that reflect intent, invoices reveal financial reality: fragmented supplier long-tail vendors, maverick spend outside the PO process, price deviations from contracts, and hidden surcharges that quietly eroded margins.

They exposed purchasing behavior, true supplier relationships, compliance gaps, and decision patterns that had previously gone unnoticed. That was when they started asking what else companies might want to do with this data and building a product around this.

“That's when we started asking what else people actually want to do with this data.” – Bella Wu

From accounts payable to procure to pay

What followed was an upstream expansion. Dexter began moving earlier in the procurement lifecycle, beyond accounts payable and into supplier evaluation, contract management and purchasing decisions themselves. 

"We realized there was so much more you could do with the data. Not just analytics, but helping people make better decisions before the invoice even arrives" Bella explains. 

Today, Dexter positions itself as an AI-first procure to pay platform. It spans the entire procurement journey, from the moment someone in a company decides they need to buy something, through supplier onboarding, contract and compliance checks, and finally PO and invoice validation and payment.

"The current product is an AI procure to pay product. We've built multiple AI agents that help throughout the whole procurement process,” says Caterina. 

Bella describes the system less as a tool and more as a connective layer across decisions that are already happening inside organizations.

"It covers everything from onboarding, managing orders, setting up and managing contracts and then validating purchase orders and invoices. It's the whole spectrum." – Caterina Kiehntopf

Why Europe is the starting point

Dexter's focus on Europe is not a marketing choice. It’s structural. Procurement in Europe comes with its own regulatory requirements, data privacy expectations, and supply chain transparency obligations that are often treated as edge cases by global software.

“We specifically focus on European companies because there are real differences in how procurement works here. There are EU-specific regulations when it comes to supply chain transparency and risk. Data privacy and where data lives really matter here," relays Caterina. 

Concerns around data residency, compliance, and reporting are of utmost importance for the companies Dexter works with. Most of Dexter's early customers sit in manufacturing and wholesale distribution, particularly mid-sized companies across sectors like food and beverage, automotive parts, and industrial goods.

"A lot of our customers are classic Mittelstand companies. They're complex enough to need proper systems, but not so big that enterprise software actually works for them." – Caterina Kiehntopf

Working together before building together

The founding team came together through shared experience at Dance, the all-in-one e-bike and moped subscription service, where they worked closely during a period of rapid growth. What was carried forward was not just operational intensity, but alignment in how they approached problems.The timing came later, as changes within Dance coincided with the accelerating rise of AI.

"I remember my first coffee chat with Bella,” recalls Caterina. “She said, 'I want to build a company.' And I remember thinking, I want to build a company too. There were structural changes happening, and it felt like our time there had come to an end.”

Bella adds that Dance was also her first role in a hardware company and the first time she experienced the complexity of managing physical supply chains. “It was the first hardware product I worked on. Managing suppliers, dealing with manufacturing issues during the pandemic, that was eye opening.”

Those experiences revealed the problems that Dexter chose to tackle.

"We want to be the go-to procure to pay tool for European companies. Especially mid-sized companies that don't want overly complex systems." – Bella Wu

Y Combinator and committing to AI properly

Y Combinator didn't introduce Dexter to AI, but it forced the team to stop treating it as optional. For Bella, applying for Y Combinator was intentional from the start. When Y Combinator announced its first-ever fall batch, the timing was tight.

“YC was the only accelerator we were interested in,” she states. “We always knew we wanted to apply. We found out about the next batch, sat down the next day, and decided to just do it.”

Inside YC, surrounded by AI-first companies, the team shifted their internal debate decisively. "Initially it was like, okay, maybe we need to do some AI,” says Caterina, “but during YC it became clear that AI had to be the centerpiece."

“We learned to make something that people want, and how to stay alive and keep going.” – Bella Wu

Overcoming uncertainty by learning faster

For Bella, one of the hardest parts of building Dexter has been living with uncertainty. “I mean, I think in general, it's like uncertainty of if you're doing the right thing, right? It's like you just don't know,” she says. Inside established companies, strategy often appears clear — “of course, the leadership team knows… we have the strategy. It all makes sense.” 

But as a founder, that illusion disappears. “When you're in the driver's seat… you're trying to make the best decision you can with the information that you have at hand, but at the end of the day, it's like you don't know, right?” Even large companies, she adds, fail because they make the wrong strategic decisions.

Rather than trying to eliminate that ambiguity, Bella and Caterina focus on learning as much as they can as quickly as they can. “The way that I feel like you mitigate uncertainty is by trying to learn as fast as possible,” she explains.

That means breaking large decisions into smaller, testable bets — “breaking them down into smaller things that you can test and, like, hypotheses that you can validate” — so you’re not “betting the company on something that you're only going to get an answer in, like, two years.”

“Our entire roadmap is based on customer feedback. We had to learn this industry from scratch and everything we know about what we're building now is coming from customers.” – Caterina Kiehntopf

Culture before scale

Dexter is still a small team of less than 10 people working together in person in Berlin, five days a week. Leadership, at this stage, is defined by transparency rather than hierarchy.

"Normally we work in the office together. In general we're very hands on and very collaborative,” explains Bella. “I really value transparency. I've mostly worked at companies where everything was shared openly, even at large companies and this is something I want to emulate at Dexter.”

As the company grows, the challenge will be carrying that openness forward.

“Even when I was working at HubSpot with thousands of employees, everything was shared openly with employees. That was a huge learning opportunity for me to just understand how companies are run, how decisions are made, how things get done.” – Bella Wu

Quickfire round

A tool you can’t live without

Bella: "I basically do all my work with Claude. Especially the project feature. I use it to collect context across deals, calls, emails. I don't know how I'd work without it."

Recharge ritual?

Bella: "I try to exercise when I can. And recently I've gotten into puzzles. It's nice to do something tactile that's not a screen."

Best advice you’ve ever received: 

Caterina: “Don't die, in reference to the business, meaning just keep going and you will succeed.”

Founder or startup you admire? 

Caterina: “Yvon Chouinard the founder of Patagonia, for his dedication to building a lasting and sustainable business. His passion for his team and rock climbing.”

Dexter is currently hiring for a Founders’ Associate Intern to work directly with Bella, Caterina and the founding team and help with their go to market strategy. If you or someone in your network is interested, then get in touch with them via LinkedIn.