WWhite Circle is hiringProduct Designer

About White Circle

White Circle builds monitoring and validation tools for LLM outputs by blocking harmful prompts, catching model abuse, and providing analytics on top. They work with companies like Lovable and are running pilots with Replit, JetBrains, and Dust.

Series A is on the horizon.

There are 16 people, mostly engineers, based in Paris: no product managers, no rigid hierarchy, just people who care deeply about what they're building. The culture is engineering-driven and a little chaotic in the best way: decisions happen fast, ideas come from anywhere, and if you have a question, you walk up to the person who knows.

The product lives in the same territory as Datadog and PostHog. But honestly, most observability tools have interfaces that only engineers can love. The team believes it's possible to make this space clean, intuitive, and even elegant. That's the challenge.

What You'll Be Responsible For

• Build the design system from scratch. This is priority number one.

• Design product features: analytics dashboards, policy configuration forms, tracing views, and settings pages. Classic B2B SaaS with web only.

• Make complex things simple. The users today are AI researchers and safety engineers, but the product is heading toward B2C. Every interface should work for both.

• Write clear task descriptions in Linear, so engineers know exactly what to build and why. Good writing matters here as much as good design.

• Join product brainstorms. You'll sit with engineers, challenge assumptions, and propose alternatives. This isn't a "receive ticket, deliver screen" role.

• Own UX copy on your screens, such as labels, tooltips, empty states, and error messages. Every word should be intentional.

What You'll Need to Succeed

• A portfolio with real UX depth: forms, dashboards, data-heavy interfaces, B2B/SaaS products. The team will look at how you think through complexity, not just how things look.

• Experience building or meaningfully contributing to a design system, not just using one.

• Basic understanding of frontend development: responsive layouts, component thinking. You don't need to code, but you should know what's realistic.

• Strong English writing. Your Linear tickets, your UI copy, your Slack messages — they all need to be clear and well-structured.

• Self-sufficiency. You take a task, dig into it, ask questions early (not after you've gone down the wrong path for a week), and come back with something thoughtful.

• Comfort with direct feedback. The team will tell you when something isn't working. They need someone who sees that as useful, not threatening.

• Genuine curiosity about the product. "What does this feature actually do for the user?" should be a question you ask before you open Figma.

Huge Plus If You Have

• Experience with AI products, developer tools, or technical infrastructure.

• Familiarity with v0.dev, Cursor, or similar AI-assisted design/dev tools.

• A habit of researching competitors, collecting references, and bringing ideas to design reviews without being asked.

• Interest in AI safety, LLM evaluation, or the broader AI tooling space.

What You'll Win

• Equity in a pre-Series A company that already has paying enterprise clients.

• CDI — a permanent contract under French law, after a 1-month paid trial.

• Salary reviewed and increased at each funding round.

• Relocation to Paris in roughly 2 weeks, as France's innovation visa program makes this straightforward.

• A two-floor office in central Paris. Attendance is flexible: show up when collaboration matters, work from home when you're in the zone.

• Direct work with the Head of Design. No PMs translating requirements, no design committees. You discuss, you design, you ship.

What's Great About This Role

Here's what's unusual about this position: you're not inheriting a design system and maintaining it. You're building one. At a company where the product is already used by some of the most technically demanding teams in AI.

You'll be the second designer. That means your thinking will shape the product, not in a "they value your input" corporate way, but in a "you'll sit in the brainstorm, disagree with the CTO, and your version might win" way. There's no product manager between you and the engineers. Features go from discussion to design to production in days, not quarters.

The problem itself keeps getting more interesting. AI outputs are getting harder to evaluate, harder to trust, harder to monitor. The interfaces that make this manageable don't exist yet. You'd be designing them.

"I think it's a rare chance to design for problems no one has really solved well yet — how do you make AI safety tools feel clear and usable? There's no playbook for that, which is exactly what makes it interesting. The team is small enough that your opinion actually changes things, and the clients are the kind you'd want in your portfolio.”

— Catalina K. former product designer, ops manager at hirehire