Security Engineer (AI Compliance & Offensive Security)

About the role

We are looking for a hands‑on Security Engineer who will own the design, implementation, and automation of security controls across AI/ML workloads, cloud environments, and offensive testing activities. You will work closely with engineering, product, legal and leadership to embed security and compliance into every layer of our platform.

Key responsibilities

• Design, build and operate security controls for AWS, Azure or GCP environments, including IAM, network, and application security.
• Develop and maintain automation pipelines (Python, Bash, PowerShell, Terraform, CI/CD) to streamline security operations.
• Conduct vulnerability assessments, penetration testing, red‑team exercises and provide remediation guidance.
• Implement and manage compliance programs for ISO 27001, SOC 2 and the DPDP Act, maintaining evidence and supporting audits.
• Secure AI/ML and LLM applications against prompt injection, data leakage and model abuse.
• Collaborate with a Virtual CISO to define security standards and mentor the growing security function.

Required profile

• Self‑driven professional comfortable in fast‑paced startup environments.
• Strong problem‑solving ability with end‑to‑end ownership of security initiatives.
• Excellent verbal and written communication skills for cross‑functional collaboration.

Required skills

• Python, Bash, PowerShell scripting.
• Infrastructure as Code with Terraform.
• CI/CD pipeline integration.
• Cloud platforms: AWS, Azure, GCP.
• IAM, application and infrastructure security.
• DevSecOps practices.
• Vulnerability Assessment and Penetration Testing (VAPT), red‑team techniques.
• AI/ML frameworks and LLM security.
• Compliance knowledge: ISO 27001, SOC 2, DPDP Act.

What we offer

• Full ownership of the security function with direct mentorship from an experienced Virtual CISO.
• Opportunity to shape security standards in a high‑impact, high‑visibility role.
• Dynamic work where building, automating, defending and attacking security tasks vary week‑to‑week.