AI Security Engineer - DevSecOps for Frontier ML

Academic Key; Cambridge; Hybrid; GBP 70,000 - 110,000; Full time

Benefits offered by this job

Hybrid work model

Travel opportunities

Job summary

CuspAI is building with a mission to secure its frontier AI platform in Cambridge, UK, focusing on end-to-end security for machine learning lifecycles and protecting proprietary models and data. The role demands security-centric thinking as we scale our research workflow and deploy guardrails across development and deployment.

You will architect shift-left security practices, integrate automated protections into CI/CD, and collaborate with AI researchers to maintain fast innovation while keeping

Qualifications

• 3+ years in a DevSecOps, Cloud Security, or SRE role in a fast-paced startup or AI-centric company.
• Proven experience building and securing CI/CD pipelines (GitHub Actions, GitLab CI, or Jenkins).
• Strong proficiency in Infrastructure as Code (Terraform, Pulumi) and securing cloud environments (AWS, GCP, or Azure).
• Deep understanding of container security and orchestration (Docker, Kubernetes).
• Practical knowledge of security frameworks and how to implement them without slowing down development cycles.

Responsibilities

• Secure end-to-end ML lifecycle and defend against data poisoning to protect proprietary models and weights.
• Design and deploy prompt injection defenses and runtime firewalls for LLM interfaces.
• Implement AI Security Posture Management (AI-SPM) to monitor self-hosted models, datasets, and endpoints.

Skills

DevSecOps

Cloud Security

CI/CD Pipelines

Infrastructure as Code

Docker/Kubernetes

Security Frameworks

Tools

Terraform

Pulumi

GitHub Actions

GitLab CI

Job description

CuspAI is building with a mission to secure its frontier AI platform in Cambridge, UK, focusing on end-to-end security for machine learning lifecycles and protecting proprietary models and data. The role demands security-centric thinking as we scale our research workflow and deploy guardrails across development and deployment.

You will architect shift-left security practices, integrate automated protections into CI/CD, and collaborate with AI researchers to maintain fast innovation while keeping