GenAI Security Architect — Hybrid Role in Portugal

IRIUM Portugal; Lisboa; Hybrid; EUR 57,000 - 61,000; Full time

Benefits offered by this job

Professional development

Flexible pay & medical insurance

Unlimited access to training

Job summary

IRIUM Portugal is seeking an AI Security Architect to safeguard Generative AI initiatives, ensuring secure architecture, governance, and validation of security controls across AI platforms and integrations.

The role emphasizes security governance over model development, requiring strong communication with Security, Engineering, Product, and Data teams in English and a track record in defending AI systems against adversarial risks.

Qualifications

• 5+ years of experience in cybersecurity domains such as application security, cloud security or security architecture.
• Strong understanding of Generative AI / LLM security risks and controls.
• Experience reviewing AI/GenAI use cases from design to production and performing architecture/security reviews.
• Ability to define security requirements for AI platforms, APIs, tools, and integrations.
• Experience with IAM, API security, secrets management, logging, monitoring, and security validation.

Responsibilities

• Conduct architecture reviews, security reviews, and threat modelling for AI/GenAI initiatives.
• Define secure-by-design guardrails and guardrails for AI platforms and workflows.
• Validate security controls, auditing, logging, and monitoring requirements across AI solutions.
• Collaborate with Security, Engineering, Product and Data teams to implement risk-aware AI architectures.

Skills

Cybersecurity

Application Security

Security Architecture

GenAI Security

Threat Modelling

IAM & API Security

Data Protection

English Proficiency

Job description

IRIUM Portugal is seeking an AI Security Architect to safeguard Generative AI initiatives, ensuring secure architecture, governance, and validation of security controls across AI platforms and integrations.

The role emphasizes security governance over model development, requiring strong communication with Security, Engineering, Product, and Data teams in English and a track record in defending AI systems against adversarial risks.