Sr. Product Designer, AWS Search, Security, Observability, Governance

Observability is being rewritten. For a decade the job was to watch servers, services, and dashboards — metrics, logs, and traces for systems that did what they were told. That world is ending. The systems people now have to understand reason, call tools, and act on their own, and they generate signal at a scale and unpredictability the old tooling was never built for. When something breaks at 2am, the question is no longer only "which host is down" it becomes "what did the system decide, why, and what did it touch." AWS is where that gets solved at scale. Hundreds of thousands of engineers depend on it to understand their applications, infrastructure, and now the AI agents running inside them. We're reimagining the whole experience for this shift — from how you explore and query telemetry, to how you manage service health and SLOs, to how you observe and debug autonomous agents, to the AI agents that help engineers investigate and resolve issues, to how customers migrate off legacy tools and land somewhere genuinely better. We're looking for a Sr. Product Designer to own meaningful surfaces of that portfolio. This is not a role that decorates engineering decisions. You'll be shaping how engineers reason about systems that are, themselves, increasingly hard to reason about. The problems are genuinely unsolved, the mental models aren't settled, and the person in this seat gets to help set them. If you want a canvas where design thinking directly shapes what gets built on one of the hardest, highest-stakes problems in cloud right now then this is it. Key job responsibilities 1. Own end-to-end design for the agent observability and ops portfolio from problem framing and IA through interaction, visual craft, and ship. 2. Invent the mental models, not just the screens. The hardest part of this domain is making invisible, high-cardinality, fast-moving system behavior legible — you'll define how engineers navigate it, what they see first, and how they move from signal to cause. 3. Design for the agentic shift from both sides: experiences that help people observe, trust, and debug AI agents — what an agent decided, the tools it called, where it went wrong — and experiences where AI agents help engineers investigate, operate, and resolve issues, with the human firmly in the loop. 4. Design across the surfaces engineers actually work in — console, IDE, CLI, and API-driven experiences — so observability meets them where the work happens instead of forcing a context switch. 5. Partner with PM and engineering from the start of each problem, not the end. On this team design has a seat in defining what to build, and you're expected to use it. 6. Run the loop from evidence — usability testing, customer sessions, and telemetry — and let what you learn reorder what you build. 7. Raise the craft bar across the area you own and the designers around you — through critique, patterns, and a point of view, contributing to a shared design system rather than one-off screens. 8. Design for adoption, not just the happy path — helping customers move off legacy observability tools and land somewhere that feels better, not just different. A day in the life Morning. You start on the problem you're closest to — say, how an engineer traces a failing request across a dozen services, or works alongside an AI agent that's triaging an incident and proposing a fix. You've got a flow in progress and a few open questions from yesterday's engineering sync. You take the first focused block to push the interaction model before the day fills up. Midday. A working session with your PM and a couple of engineers. Not a hand-off — a real debate about what the primary view should lead with and what can wait a layer in. You're arguing from the customer's reasoning path, they're arguing from what's queryable in real time and what's cheap to compute, and the design gets sharper because both are in the room. Afternoon. You sit in on a customer usability session. You watch someone get lost exactly where you suspected they would — and somewhere you didn't. You come out with a list that reorders your week. Later, a design critique with the broader team: you bring your flow, get it pressure-tested, and give sharp feedback on someone else's work. The throughline. You're never far from a real customer problem, you're designing in a space where the right answer isn't known yet, and the work you do today shows up in the product in weeks, not quarters. The pace is real — but so is the ownership. About the team We're the AWS Security, Search, Observability & Governance design team. Our mandate is to make the most technical, highest-stakes workflows in cloud feel clear, fast, and even good to use. What's distinctive about us: - Design is a competitive weapon here, not a line item. We're betting that in a market still figuring out what observability for the AI era even looks like, the team with the clearest experience wins. Leadership backs that bet with real scope for designers. - We ship, and we ship hard things. New exploration experiences, agent observability, migration tooling, core monitoring at massive scale — these are launching now, to real customers, against real competitors. Your work isn't theoretical. - We work from evidence. Usability labs, customer sessions, and telemetry are part of the rhythm, not an afterthought — so your point of view is grounded, and you get the satisfaction of watching it land. - The bar is high and the people are senior. The designers around you have deep backgrounds in AI, cloud infrastructure, and enterprise platforms. You'll be challenged, and you'll have peers worth learning from. If you want to do the defining design work in observability for the AI era on a team that will actually let you come build it with us.