Software Engineer, AI Security

We are building a first-of-its-kind vulnerability auditing and review platform for LLM and foundational model integrated software — including components like foundational models, MCP servers, agents, and agent skills/capabilities. As an SDE 1 on this team, you will contribute to the design, development, testing, deployment, and operation of security-focused software systems that help identify and mitigate vulnerabilities across AI-integrated applications. This is an opportunity to work at the intersection of AI safety, security research, and platform engineering within one of Amazon's most innovative organizations.

- Design, implement, test, deploy, and maintain security-focused features for the vulnerability auditing and review platform.

- Build and extend vulnerability detection capabilities for LLM-integrated services and tools including foundational models, MCP servers, agents, and agent skills/capabilities.

- Deliver working software features end-to-end, working backwards from customer requirements through production deployment and ongoing operations.

- Refactor and deprecate existing systems as the platform evolves, ensuring long-term maintainability and extensibility.

- Develop automated scanning and analysis tooling to identify security vulnerabilities in AI-integrated software components.

- Implement detection mechanisms for adversarial attacks, prompt injection, capability boundary violations, and other LLM-specific threat vectors.

- Collaborate with applied scientists to translate AI safety research into production-grade platform features.

- Apply secure coding practices and threat modeling to all platform components.

- Own operational health of your features — monitor, alarm, triage, and resolve production issues.

- Identify root causes of operational failures and implement permanent fixes; never settle for the status quo.

- Proactively identify and execute on opportunities to improve the team's operational posture (automation, runbooks, telemetry, dashboards).

- Make effective priority tradeoffs between new feature development and operational improvements.

- Design software solutions that bring clarity to difficult problems with visible risks or roadblocks.

- Participate actively in code reviews, providing meaningful and constructive feedback to engineers at all levels.

- Document how the platform is constructed, tested, operated, and secured — and how it fits into the broader AI safety ecosystem.

- Keep skills current; evaluate and apply industry innovations in security, AI safety, and software engineering where applicable.

- Actively mentor SDE 1s and new team members to accelerate their productivity and growth.

- Train new teammates on platform architecture, operational practices, and team norms.

- Contribute to hiring efforts through interviews, candidate assessments, and bar-raising.

- Work with customers, stakeholders, and peers (including applied scientists) to understand business and customer value and ensure the platform solves the right problems.

- Resolve disagreements with peers through constructive, inclusive dialogue.

- Communicate effectively about the work you deliver — status, risks, tradeoffs, and outcomes.