Security Solution Architect, APJC Customer Security

Amazon Web Services (AWS) is seeking a Secure AI Solution Architect to join our APJC Customer Security team in a content-and-frameworks role. Customers across the region are asking the same questions about building securely with AI — how to give autonomous agents an identity, how to defend against prompt injection, how to govern the AI development lifecycle, how to evidence AI governance for a regulator. The approaches that answer these questions are being proved with customers today; what's missing is the layer that turns them into clean, documented, deployable patterns the whole field can use. You build that layer. Working hands-on, you take validated approaches and turn them into reference implementations that work out of the box. You test them against real customer requirements, then package them into a published library that Solutions Architects across the region draw on. Your work is the difference between a pattern proved once and a pattern the whole region can reuse. You are technically self-sufficient: you write the code, build the reference implementation, test it, and document it so a Solutions Architect in any area can adopt it without specialist presence. We are looking for a builder who enjoys taking a defined but difficult technical problem and turning it into an elegant, well-documented answer others can build on. Deep AWS experience helps but is not required at entry; we can teach the services. What we need is strong hands-on technical craft, security depth, and the instinct to build things that scale beyond yourself. Key job responsibilities - Take validated Secure AI approaches (agentic-AI identity patterns, AI security frameworks, AI development lifecycle controls, guardrail and governance-evidence patterns) and build them into clean, deployable reference implementations — working code, not slideware. - Test each pattern against real customer requirements to confirm it holds, then package it into the published library that the field draws on. - Write the code, reference architectures, and documentation to a standard that a Solutions Architect or partner in any APJC area can adopt without specialist presence. - Work with the AWS service teams (Bedrock, AgentCore, SageMaker) — bringing field requirements and customer data into the product conversation and taking away clarity to turn into guidance. - Stay current on the AWS AI and security product set, translating new and upcoming capability into field-ready reference implementations as it lands. - Surface customer needs and gaps as feature requests, contributing to the feedback loop that shapes what ships next. A day in the life You start the morning building out an agentic-identity reference implementation that two areas have asked for. A customer engagement last week surfaced a gap around cross-account tool-use authorisation; you build the fix, test it, and update the library entry so every area gets the improved version. Mid-morning you join a session with the AgentCore service team. You bring the field data — what APJC customers are hitting on non-human identity — and take away clarity on what can ship as guidance now versus what needs a product change. You'll turn that into an updated reference architecture this week. After lunch you go heads-down and build — hands on keyboard — turning an AI development lifecycle control you prototyped into a documented reference implementation with code, deployment steps, and a short workshop a field Solutions Architect could run themselves. You close the day handing the finished pattern to the go-to-market team to fold into the next immersion-day curriculum, publishing the write-up to the field, and briefing the Associate SA on what's ready to be packaged into next week's community event. No two days look the same, but the thread is consistent: you build it once, and the whole region gets to build on it.

APJC Customer Security helps customers across the region adopt AI on AWS with