Security Engineer

About the role

The Security Engineer designs, implements, and operates security controls across identity, endpoints, networks, cloud services, applications, data, and AI-enabled systems. This hands-on role reduces cyber risk while enabling the responsible use of modern technology. The engineer partners with infrastructure, application, data, legal, privacy, and business teams to build resilient services, detect and respond to threats, and translate technical findings into practical action.

Responsibilities

• Engineer secure platforms. Design and improve preventive, detective, and recovery controls across hybrid infrastructure, SaaS, cloud platforms, endpoints, networks, and business applications.
• Strengthen identity. Implement MFA, Conditional Access, least privilege, privileged-access management, lifecycle governance, service-identity controls, and periodic access reviews.
• Secure cloud-native workloads. Assess cloud configurations, APIs, containers, serverless services, infrastructure as code, CI/CD pipelines, secrets, and software supply-chain dependencies.
• Manage exposure. Identify and prioritize vulnerabilities, attack paths, misconfigurations, unsupported assets, and internet-facing risk; partner with owners through verified remediation.
• Embed security by design. Perform architecture reviews and threat modeling for new systems, integrations, vendors, and major changes; define proportionate control requirements before production.
• Govern enterprise AI use. Maintain visibility into approved and unapproved AI services, models, agents, copilots, plugins, data connections, and business use cases; help teams choose secure, sanctioned paths.
• Secure AI systems end to end. Review data flows, model and API access, retrieval pipelines, vector stores, prompts, tools, memory, outputs, and human-approval points across the AI lifecycle.
• Threat-model AI-specific abuse. Evaluate prompt injection, sensitive-data disclosure, insecure output handling, model or data poisoning, excessive agency, identity abuse, supply-chain compromise, denial of service, and unsafe tool execution.
• Implement AI guardrails. Apply strong authentication, scoped authorization, data classification and DLP, secrets management, content filtering, tool isolation, rate limits, audit logging, monitoring, and human-in-the-loop controls.
• Validate before and after launch. Coordinate security testing, adversarial evaluation, red teaming, misuse-case testing, and ongoing monitoring for AI