Want to know if this job is worth applying to?
Tallinn, Eesti
Hybrid
Katana builds powerful inventory and manufacturing software for 1,500+ businesses that make and sell physical products around the world. We're an AI-first SaaS company, and that shows up in how we build as much as in what we build.
We've just stepped into an exciting new phase of expansion - moving beyond small merchants into a larger market of modern, multi-channel, multi-location businesses. With that growth comes bigger opportunities, more complex problems, and the chance to shape how we scale.
Our 100+ person team across 10+ countries is building what's next - together. And that's where you come in.
Three things about this seat worth knowing before anything else.
Real ownership. You own a product domain end to end - the customer interviews, the problem framing, the design, the thing that ships, and the number that moves afterwards. Roadmap items at Katana aren't a product manager's by default. Designers and engineers lead them too, and we'll expect you to want one.
Speed. Sketch an idea Monday, put a working prototype in front of a real customer by Friday - built with AI on our design system, running on their own data. We work in cycles of two weeks or less, in vertical slices. Designers here have already taken roadmap work from prototype straight to a production pull request, and making that the default is part of this job.
Visibility. You report to the CPO. You present your own work, your own experiments, and your own misses - directly, to the person making the calls. Nothing gets absorbed into team output on its way up.
This isn't a role for someone who wants a bigger version of the job they already have. It's for someone ready to own something that matters, at a company moving fast enough that your work shows up in the product within weeks, not years. If you want to shape how a product gets designed while that is still being worked out, this is a seat where you actually can.
This role is hybrid from our Tallinn office. Tuesday to Thursday are office days; Monday and Friday are up to you. It's the way our product team works - that's when the whole team is in the room, and the face-to-face time makes the work move faster and the collaboration sharper.
If you're not already in Estonia, we're happy to support relocation from within Europe for the right person.
You'll join the Platform & Intelligence team - the team that pushes forward every AI initiative at Katana - and you'll report directly to our CPO. You'll work in a pair with another senior product designer, splitting the ground between you and debating the hard parts together.
It's a senior seat with real scope, and it sits in the most interesting place in the company right now.
This is frontier work in the most literal sense. We're building towards a Katana that a customer can simply talk to: ask "did I actually make money on this product?" in plain English and get a trustworthy, sourced answer out of their own operational data, then act on it - creating real purchase orders and manufacturing orders in a live system of record. We're also putting Katana inside the AI tools customers already use, through connectors for Claude, ChatGPT and Gemini, so we own the experience without owning the screen it renders on.
Nobody has a finished design language for any of this. How you show a customer where an answer came from so they trust it. How you ask for confirmation before writing to live inventory, and how you let someone undo it. How you make an assistant saying "I can't answer that one yet" read as confidence rather than failure. What a design system even needs to become when a growing part of the product is conversational rather than composed of screens. These are open design problems, and they'd be yours.
To be clear about scope: this isn't a design-ops or pixel-execution seat. We want the designer who framed the problem, not the one who styled the answer. You'll do your own research, argue for the user in roadmap decisions, and be accountable for whether the design moved a number. The learning curve is real - Katana is a deep operational product and our customers' businesses are genuinely complicated. So is the opportunity.
This is the main focus of the role, the team's core bet, and the reason the seat exists.
Own the design of the in-product Katana AI experience end to end, from the first prototype through beta to something customers rely on daily.
Design the trust layer, which is the hard part: sourced answers, confirm-before-write safeguards, permissions, audit trails, and undo for anything the AI changed.
Design for the moment the assistant doesn't just report a number but explains why it's wrong and offers to fix it. Our strongest customer signal is that people don't trust the figures they see. Solving that is the differentiator.
Shape how proactive insight reaches customers, so they hear about a stockout, a late supplier, or a margin slip before it hurts them, without the product turning into noise.
Work on how Katana shows up inside other AI tools through our MCP connectors, where we shape the experience without controlling the interface.
Run the research that tells us which of this is real, and set the KPIs that tell us whether it worked.
The AI work is the job. But we're a small design team and we cover for each other, so you'll also pick up work outside it when it's needed. Two places that's likely to happen - neither of these is the core of the role, but both are real, and we'd rather tell you now than surprise you later.
Advantage. The Advantage team builds custom solutions against the Katana API for our highest-value SMB customers, with AI doing much of the building. They'll sometimes need design input on short-timeline work where the requirement is specific and the customer is real. The challenge is solving one customer's problem properly without quietly breaking the product for everyone else - and custom builds are one of the sharpest signals we have about where Katana falls short, so what you learn there should feed back into the core product.
The design system. Nobody currently owns it, and we'd love someone to want to. Understand what it's really worth here: it isn't a library designers browse for components. It's what AI builds against, which makes it the highest-leverage artefact in the whole product - every decision inside it gets multiplied across everything we build. We'll be honest: the product isn't as consistent as it should be yet. Deciding where consistency genuinely serves the customer and where it's just habit, then changing it, is part of the work for whoever picks this up.
Design at Katana is built, not drawn. You prototype with AI against our own design system, so what you put in front of a customer is a working thing running on their real data, not a picture of one. The feedback you get on that is sharper than any mockup review, and it's something we couldn't do before. Designers also have direct access to product and customer data - you don't queue behind a PM or an engineer to answer your own questions.
Where this is heading: design that goes from prototype straight to a production pull request, with no handoff in between. That isn't the default yet, but it has already happened here - designers on the team have shipped roadmap work this way - and the process is being built in the open. Part of this job is helping make it the norm. We won't expect you to arrive knowing our stack. We will expect you to want to work this way.
Prototype with AI against the Katana design system, on real data, and get it in front of customers early.
Take work through to production yourself where you can, and help close the gap where we can't yet.
Deliver in increments that add up to a coherent experience, rather than one big reveal.
Work shoulder to shoulder with engineers, product managers, marketing and the customer-facing teams. Building your own work doesn't mean working alone, and the engineers on your team are the people who'll get you good at this fastest.
Know the roadmap, contribute to planning, and understand how the product works at a level that lets you argue with it.
Run customer interviews yourself. Not a summary of someone else's.
Use behavioural tooling (FullStory or similar) and research documentation (Dovetail or similar) to track what customers do and keep what you learn where the team can use it.
Apply real analysis to what comes back, and make decisions you can trace to evidence.
Keep learning the industry and the specific problems Katana customers face, and bring that back into the product.
Regularly revisit shipped features and push them from MVP to MLP - from merely viable to something customers actually like using.
Focus on engagement and satisfaction, not just launch.
Set the KPIs for your own work up front, and evaluate design changes against them.
Gather qualitative feedback alongside the numbers.
Present outcomes to the team - what worked, what didn't, and what you'd do differently - so the rest of us get smarter too.
First 30 days: You're not warming up on the sidelines. You've gone deep on the product, our customers, and where the AI work currently stands, and you've already built your first working prototype with AI on our design system and put it in front of a real customer. You and your design partner have a clear split of the ground between you. You're asking sharp questions about the parts of our process that are still being figured out.
By 6 months: You've designed AI work that real customers use, and you can explain what the research told you and where it changed your mind. You're running the full loop end to end - research, design, build, ship, measure - and you've taken at least one piece of work all the way to production yourself. The Advantage team knows to bring you in early rather than late.
By 12 months: Katana AI is visibly better designed than the day you arrived, including things you went back and rebuilt from viable to lovable. The patterns you've set for trust, confirmation and explanation are how Katana does AI. Design going straight from prototype to production is closer to the default than the exception, and you had a hand in that. The way you work - how you research, how you set KPIs, how you present a miss - has become part of how the whole team works.
You have 5+ years as a product designer on B2B SaaS products, with a track record you can walk through - what you owned, what you shipped, and what happened next.
You want to build the thing, not hand it over. Maybe you already design with AI tools, maybe you write front-end code, maybe you've simply never accepted that your job ends at a handoff. We care that you want to own the built result. We don't care which route got you there, and we'll teach you ours.
You run user research yourself - interviews, behavioural tracking, documented insight - and you analyse it properly, using more than one method. You can point to a decision that changed because of what you found, not one that was decorated with it afterwards.
Strong UX and UI craft. You know the difference between the two and you're good at both, and when a design is nearly right you can explain exactly what's wrong with it and what it costs the person using it.
You know when to work inside a design system and when a problem genuinely needs a blank canvas - and you've done both. You care about consistency and know how to move a product towards it, rather than expecting to inherit it.
You can design where there's no established pattern to copy. Much of the AI work has none, and you should find that energising rather than unsettling.
You ship a v1 that's good enough to learn from, then make it better. You've taken an existing product further, and you've set the KPIs for that work yourself rather than being handed them.
You carry a business mindset. Bring an example where you can state both the customer problem and the business problem you were solving.
You've designed an AI or conversational product surface, and have opinions about what works and what doesn't.
You already build with AI coding tools, or you're a designer who has shipped their own front-end code.
You've owned or meaningfully shaped a design system.
You've used the Jobs To Be Done framework on real work, not just read about it.
A formal visual or graphic design education.
You know the operational world - inventory, supply chain, manufacturing, or ecommerce operations - well enough to follow a customer's first explanation of their setup.
An owner, not an order-taker. You take responsibility for your domain, know the roadmap, and contribute to planning rather than waiting for it.
Independent and collaborative at the same time. You can run alone, and you know great products aren't built that way. Designing in a pair should sound better to you than designing solo.
You share early and often - rough work in front of designers, PMs, engineers and the customer-facing teams, not a reveal at the end.
Flexible about method. You know many ways to get to a result and accept that the right one depends on the circumstances. We change often and fast, and that should energise you rather than scare you.
AI-first for real. This isn't a company that encourages you to try a chatbot now and then. AI has already changed what a designer does here day to day, and working out what it changes next is part of the job.
Growth mindset. You try different ways of getting to impact, experiment with methods that might not work, and present your failures to the team so others can learn from them.
You're genuinely interested in how customers run their businesses, and you're willing to learn a deep operational product properly. Katana is complicated and that should appeal to you, not put you off.
A competitive, market-aligned base salary
Employee Stock Option Program - you share in what you build
Time off that's genuinely encouraged: statutory annual leave aligned with local requirements, with a global minimum of 4 weeks, 3 paid health days annually, 4 days of time off in lieu, and your birthday off.
Fully paid 1-month sabbatical + 1,000 EUR travel budget after every four years at Katana - time to reset, explore, or focus on personal growth
Home office stipend to help you create a setup that works for you
Health & wellness package: regular health checks through Confido, plus a yearly budget you can put towards Confido health insurance, Stebby sports and wellness compensation, or a mix of both
Modern tools and tech, including AI-enabled tools that help you work smarter and stay focused on high-impact work
Automated Screening & AI Disclosure: We use automated tools including AI to help review applications and filter candidates. Final decisions are always made by people. If you're in a place that requires it, we'll ask your consent. Ask us anytime about what this means.
