User Experience Researcher

Anthropic•San Francisco, NY

4d•Hybrid

About The Position

Anthropic's mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems. The UX research team at Anthropic is a small team with outsized leverage. We work across products used by millions of people—and on interactions that don't have established conventions yet. We inform critical decisions through rigorous research, with stakeholders who eagerly incorporate findings into their work. Research has a seat at the table for product strategy, and we partner closely with design, product, and engineering. We combine qualitative and quantitative methods to understand user needs in a space that's evolving faster than anyone can track. We move fast, favor impact over polish, and increasingly build our own tools—often with AI doing the heavy lifting—to do research at the speed this environment demands. This is a senior individual-contributor role. You'll own research for a complex, fast-moving product area where the surface is expanding quickly and the most important user questions haven't been articulated yet, let alone answered. You won't work from a fixed research roadmap. You'll prioritize constantly—making clear, well-informed bets about what to learn next, executing on them fast, and pivoting the moment the picture changes. Our product teams are deeply technical and move on instinct; your job is to make that instinct better-calibrated without slowing it down. Roles here are fluid. On any given week you might be the researcher, the analyst, the one framing the next product bet, or the one building a prototype because that's the fastest way to find out. Where this role lands depends on team needs and your strengths—we'll figure that out together.

Requirements

Industry UX research experience, including ownership of research strategy for a product or product area
Track record of research that shifted product direction at the strategy level, not just the feature level
Deep mixed-methods expertise—qualitative and quantitative
Enough data fluency to use Claude to work directly with our data warehouse—pulling your own samples, writing your own queries, running your own analyses
Demonstrated technical range: you've built research tooling, prototyped, scripted analyses, or otherwise shown you can keep up with very technical teams without translation overhead
Advanced, hands-on use of AI in your own research workflow—you're already taking full advantage of its capabilities to enhance how you work
Experience embedding with engineering-led product teams that ship fast, and earning influence in that environment
Excellent storytelling, calibrated to audience. You can discuss a finding with an exec as effectively as with an engineer

Responsibilities

Set the research agenda for a product area where the strategic questions are still fuzzy. Define what we need to learn, in what order, and why—then go learn it. Don't wait to be asked
Design and run mixed-methods research at whatever fidelity the moment calls for. Some of it will be careful, foundational work the team references for a year. Some of it will be a finding you ship the same day the question came up
Keep product teams close to users in a rigorous way—building the channels, cadence, and evidence base that make user reality hard to ignore
Get hands-on with product data. Pull your own samples, build your own dashboards, instrument what's missing
Prototype when prototyping is the fastest path to an answer—a research tool, a survey pipeline, a scrappy version of a feature to put in front of users
Deeply leverage AI in how you work—using Claude to make yourself dramatically more efficient at analysis, synthesis, tooling, and everything in between
Use what you study. Dogfood our products hard enough to develop your own informed opinions, and bring that texture into your research
Anticipate where the product is heading before anyone asks. Show up with the user understanding the team is about to need
Turn what you learn into recommendations confident enough to redirect a roadmap—including recommending we kill something that isn't working
Build durable understanding, not just findings. The goal is that PMs, designers, and engineers internalize the user mental model well enough to make good calls when you're not in the room
Develop new research methods when the existing playbook doesn't fit—particularly methods that use AI to do things that weren't possible before