Steward is building AI infrastructure for complex AML and entity onboarding.
Outward recently sat down with Steward co-founders Arik Oslerne (Co-founder & CEO) and Moshe Lieberman (Co-founder & CTO) to talk about the experiences that shaped them, how they found each other as co-founders, and why they believe AI can finally solve one of financial services’ most painful operational problems.
From judo mats and entrepreneurial family roots to first-principles company building, Steward’s story is one of speed, standards and sharp conviction.
1. What personal experiences have shaped the course of your lives?
Moshe
My dream was to become an Olympic judo champion, and I was somewhere on that path – training six days a week, sometimes twice a day. Beyond the competitiveness and methods, this influenced how I approach building. The work ethic and the ability to persevere through difficulties and challenges. I don’t think I’d be the same person without it.
I was also always curious about tech, but I didn’t really have time for it until I got injured. Suddenly I had all this free time, so I started teaching myself to code. Looking back, judo gave me the mentality, and programming gave me the next direction.
Arik
My childhood was shaped a lot by my father. He was very entrepreneurial. He moved from Israel to the UK in the early eighties, barely spoke English, and decided to build a business anyway. He’d drive across the country reselling machinery.
At the time, I used to get annoyed because every time I got in the car with him, he’d immediately call someone about business. I’d think, where’s my time with my dad? But looking back, it was probably the best MBA I could have had. Listening to those negotiations, the relationship building, the instinct for opportunity, all of that shaped how I think about business and people. It also made partnership feel very natural to me. When I started Steward, having a co-founder wasn’t optional. It was critical.
2. How did you land on the idea of Steward and why does the world need it?
Arik
AML has been a pain point in every part of my career – fund set up, brokering, private wealth management. What brought it home was opening a private wealth account after my exit from Vauban. The process dragged on for nine weeks. I sent documents, then more documents, then filled out forms repeating what I’d already provided, then waited while it went back and forth with compliance.
At the end, I remember calling my account manager and saying: “Seriously, what’s going on? Do you think I am a drug lord? A terrorist? Just tell me.” And he said “no, no, sorry, it’s just with compliance”.
That moment made it very obvious to me that the amount of time and human effort being spent on this was absurd. Someone had to solve it.
Moshe
The most important thing was how universal the pain is. Every single conversation was the same reaction. Immediate recognition, visible frustration. That told us very quickly we were in the right place.
3. Why is this problem solvable now in a way it wasn’t before?
Moshe
Two years ago, you could solve a narrow use case: one document type, one fairly standard structure. Step outside those boundaries and the results fell apart. For complex onboardings with messy structures, sprawling org charts, inconsistent documentation, the AI just wasn’t good enough before. Especially for what comes after first-line checks – review and audit. The steps that require actual out-of-the-box critical thinking.
That’s changed. The LLMs are improving fast, and that means we can keep expanding the range of use cases, the speed, and the accuracy. Without that shift, we simply wouldn’t be able to build this product.
4. How did you meet, and how did you know you were the right fit as co-founders?
Arik
I knew I wanted to build Steward with someone. Startups are hard. You want to go through that journey with someone who can push you forward, pick you up when needed and challenge you when necessary.
I was getting Shai Wininger’s (founder of Lemonade) thoughts on the idea for Steward, and at the end I asked if he knew anyone who’d be the right fit for a co-founder and CTO. He immediately mentioned Moshe.
We spoke on Zoom that same day. Very quickly, everything clicked: values, vision, how we wanted to build, even the things we didn’t want. Within a week I was on a flight to New York and spent an intense week together digging into the problem, the market and the product. Then it became about pressure-testing it with customers to know we were on the right track. Things moved quickly from there.
Moshe
Before I met Arik, I’d spent a while trying to find the right co-founder and the right problem. I’d even started something that didn’t work out. My big takeaway was that if I wasn’t the domain expert, the person next to me had to be. Arik fit that bill.
We also balance well. If we had to choose between quality of output, time it takes to do a task, or the cost of doing it, Arik would naturally optimise for time. He has a true getting-****-done mentality and operates at a pace that impressed me from the first conversation and still does. I care deeper about quality so I push back on going all out just for the sake of speed.
5. What was the most impactful customer insight to date?
Arik
There was one early customer who loved the product and kept showing it to every single person they knew. One of the things they said was: it would be amazing if we could just dump all the data into the system with no structure and let it figure things out.
Moshe took that away and built what became one of the most powerful parts of the product: bulk upload. You throw in a zip file full of documents, the system searches for the relevant ones, places them into the flow, and tells you what’s missing or what isn’t good enough. That resonated immediately because that’s exactly how these processes often work in reality, just manually.
6. What could Steward become in the future?
Arik
The most interesting part of the product for me is the idea of a Steward profile, a reusable AML and KYC profile that stays updated over time.
If you think about investing, you don’t do it once. You invest across multiple funds, managers and platforms, then there are periodic reviews, trigger events, changes in circumstance. Every single time you’re asked who you are all over again and things get very messy very quickly.
The vision is to solve that at the root. Create a standardised, reusable profile of an individual or entity, a central repository that gives a firm what it needs to make fast, informed decisions. New foundational infrastructure for the industry.
Governments are trying to build answers to that by creating repositories, but slowly and they end up very isolated. Steward will move faster. Over time, I can imagine a world where being asked “Do you have a Steward profile?” becomes normal.
7. What makes someone the right fit for your wider team?
Moshe
We want people who understand it’s not a nine-to-five job. You have to join because you’re excited about building from scratch and because the mission means something to you.
The other big thing is how AI-native you are beyond using it for coding – in sales, marketing, operations, product, everywhere. We want people who are genuinely obsessed with probing new AI tools. Not just people who obsessively use ChatGPT, not just because it’s in your job description. That makes a huge difference to efficiency, productivity and how we function as a small team, an approach we want to scale.
Arik
We focus on hiring only A-players and care about quality, accountability and intensity. Both Moshe and I have extremely high standards. In early-stage companies, getting the right person in the door can mean the difference between a 20% output and a 150% output. That’s really powerful when you’re at the zero-to-one stage.
My standard is that a customer or prospect hears back from us within six minutes. People are often shocked by that. Some then decide it’s too much, which is great, because they’ve self-selected out. We want everyone who joins to understand the level of care and urgency we expect.
8. What principles from your earlier company-building experiences have you brought into Steward?
Arik
At Vauban we went through a major pivot. As we were fundraising we had to be honest with ourselves: this could be a good business, but it wasn’t going to be a unicorn-scale business. So we cut it and pivoted hard.
It’s very easy to get trapped in the weeds of building, growing, building, growing. But sometimes the highest-leverage thing is to zoom out and ask: what do we need to do to get to the next stage, and what’s really holding us back? That means making very difficult decisions.
Moshe
Focus is everything. Even if everyone is working incredibly hard, output is still finite. So the question is whether you’re spending time and focus on the things that actually matter. There are lots of things across marketing or product you can do that seem valuable, but have very little effect on converting new customers, retaining them, or growing the business.
That focus also changes constantly. Something that looked like the highest priority at the start of the week may no longer be the highest priority after a few customer conversations. So you have to keep reevaluating.
9. What’s the best advice you’ve received from other founders?
Arik
Don’t over-index on advice. Opinions you get from your investors, team or advisors might be conflicting and you end up being pulled in different directions every day. You have to listen, analyse, question and absorb what’s useful, but ultimately you need to have conviction about what you’re doing.
Moshe
One thing I’ve learned is that it’s often better to start moving before you have all the information, and then learning as you go. I used to feel more comfortable waiting until I had a fuller picture before making a decision. But very often, it’s better to do something, even if it’s not perfect, and then iterate. The momentum itself teaches you things you wouldn’t have learned by waiting.