Security Architect (Cloud / Devsecops / Ai)

Decipher Bureau; Sydney; On-site; AUD 216,000 - 264,000; Full time

Benefits offered by this job

Sydney office

Hybrid work: 3 days in office

Job summary

Decipher Bureau is seeking a Sydney-based Security Architect with strong DevSecOps background and recent AI experience to drive security across AI agents and platform changes. You’ll threat model, design mitigations, and translate controls into code with testable pipelines.

This engineering-led role blends architecture with hands-on security work, focusing on data privacy, logging, guardrails, and runtime screening within modern CI/CD environments.

Qualifications

• DevSecOps experience across CI/CD pipelines and software supply chain security.
• Engineering mindset to read and challenge how a control is implemented in code.
• Cloud security expertise across AWS, Azure or GCP.
• Genuine interest in AI security concepts such as prompt injection and agent permissions.

Responsibilities

• Threat model AI agents and platform changes with engineering teams.
• Apply security lens to agent platform changes, including new integrations and features.
• Secure CI/CD pipelines and software supply chain across GitHub Enterprise and GitLab.
• Define pipeline tests to prove security controls work in practice, not just in theory.
• Guide engineers on guardrails built into agent frameworks and runtime screening.

Skills

DevSecOps experience

Engineering mindset

Cloud security (AWS/Azure/GCP)

AI security concepts (prompt injection

Tools

GitHub Enterprise

GitLab

Google Cloud ADK

Model Armor

Job description

Security Architect (Cloud / Devsecops / Ai)

This is a chance to get in early on one of the most interesting problems in security right now: AI agents moving from experiment into production, at enterprise scale.

Are you a Sydney based Security Architect with a strong DevSecops background and some recent experience with AI?

The organisation is large, global and well invested in AI. Agents are being built and deployed through the same pipelines as the rest of its software, and the agent platform changes constantly with new features, integrations and connectors.

They need a security architect who can keep pace with that and work shoulder to shoulder with skilled engineers.

The role sits between architecture and engineering. You will threat model new agents and platform changes, recommend mitigations, and then go a level deeper: how the control is actually implemented in code, and which tests in the pipeline prove it works.

This is an engineering-led role with architectural influence, not a governance or paper checkbox position.

What the role covers

• Threat modelling AI agents with engineering teams: data privacy, logging, human in the loop, and what systems an agent can reach
• Applying a security lens to changes on the agent platform: new integrations, connectors and features
• Securing CI/CD pipelines and the software supply chain across GitHub Enterprise and GitLab
• Defining pipeline test cases that prove security controls work, rather than assuming they do
• Guiding engineers on guardrails built into agent frameworks and runtime screening

What you will bring

• Strong DevSecOps experience: pipelines, Git platforms, supply chain security and vulnerability management, from hands-on work
• An engineering background or mindset: you understand the software development lifecycle well enough to read how a control is implemented and challenge it
• Cloud and container security across AWS, Azure or GCP
• A genuine interest in AI security, with a solid grasp of the concepts: prompt injection, agent permissions, guardrails. You don't need to be an AI engineer.
• Google Cloud agent tooling such as ADK or Model Armor (nice to have)

This would suit a DevSecOps engineer or platform security specialist ready to step into architecture, or a hands-on architect who still enjoys getting into the detail.

Salary: Up to $240,000 + Bonus

Permanent role, Sydney based, 3 days in the office.