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

Security is one of the hardest problems in the world and it just got harder. As AI agents begin to write code, triage infrastructure, and take action autonomously on customers" behalf, the old model of security - dashboards a human reads, alerts a human triages. The hard question is no longer can a person see what happened? It"s can a customer trust a system that acts on its own, verify what it did, and stay in control? That question of trust is the design problem of this decade, and it sits at the center of this role.

We"re looking for a Senior Product Designer to help reimagine what security looks like in an agentic world. You"ll own the end-to-end experience for agentic security surfaces including AWS Continuum and AWS Security Hub products we"re rebuilding from first principles around autonomy, action, and trust rather than around tables of findings. This is net-new territory: there is no established pattern language for how a customer grants an agent autonomy, watches it work, understands its reasoning, and reins it back when needed. You"ll invent it. Your work will decide whether customers feel confident handing real consequence to a system that acts for them, the difference between security that customers tolerate and security they trust.

Key job responsibilities

1. Own the end-to-end design for agentic security surfaces - AWS Continuum, AWS Security Hub, and related products - from problem framing and research through interaction design, visual craft, and launch.

Define how trust is earned in the interface: how a customer grants autonomy, observes an agent"s actions and reasoning, verifies outcomes, and intervenes - building confidence through evidence rather than demanding it up front.

2. Set the design vision and the pattern language for autonomy in security, in a space where those patterns don"t yet exist, and bring partners along with sharp narratives and prototypes.

3. Partner daily with product, engineering, and science to balance customer needs, deep technical constraints, and the novel safety and verification demands of agentic systems.

4. Confront the runtime verification problem head-on: when an agent acts faster and more opaquely than a human can follow, design the experience that keeps the customer informed, in control, and able to trust what the system did.

5. Ground every decision in evidence - synthesize qualitative and quantitative signal, run research, and validate with real security practitioners before and after launch.

6. Translate dense security and agentic concepts (autonomy levels, agent reasoning, remediation, verification, telemetry) into interfaces that reduce cognitive load and build confidence.

7. Raise the design bar across the team through critique, mentorship, and reusable patterns, and extend the shared systems your work depends on.

A day in the life

Your morning might start in a working session with your PM and tech lead, pressure-testing how much autonomy a customer should be able to grant an agent on day one - and how the interface earns the right to ask for more. Midday you"re in a critique, giving and getting sharp, specific feedback on how an agent surfaces its reasoning without burying the customer in detail. In the afternoon you watch a security analyst work through your Continuum prototype - the moment they hesitate before letting the agent act tells you more than any survey - and you redesign that moment before you log off. Across the week you move between zoomed-out strategy (what does security even look like when agents do most of the work?) and zoomed-in craft (is this the right default, the right disclosure, the right word for "undo"?). You"re trusted to set your own direction, surrounded by partners who treat the trust problem as seriously as you do.

About the team

AWS Security sits inside SSOG, where the thesis is direct: as AI systems grow more capable and more autonomous, security has to keep pace - and the gap between what an agent can do and what a human can verify is one of the defining problems of the field. We"re not bolting AI onto yesterday"s security console; we"re rebuilding the experience for a world where agents act and humans need to trust, verify, and stay in control. Our design org is the MOAT proving the clarity and trustworthiness of an experience is as much a differentiator as any capability. We"re a senior, tight-knit group that ships real products for real customers, invests in craft and research, gives designers genuine ownership of their problem space, and cares about raising each other"s work.