Principal Application & AI Security Engineer

Principal Application & AI Security Engineer

DNV Energy Systems' Platform Services is seeking a Principal Application & AI Security Engineer. DNV Energy Systems' Platform Services runs the software products and digital platforms our customers depend on, including systems with significant operational importance in enterprise and energy environments. As we evolve toward agentic AI architectures, security must move from after-the-fact review into architecture, development workflows, and runtime operations - engineered into the platform and the delivery pipeline, with evidence that controls are implemented and operating effectively.

This is a builder's role for a senior technical leader who can read and improve code, design reusable controls, model complex threats, conduct authorized security testing, and work directly with engineering teams to ship durable fixes. The goal is not simply to identify vulnerabilities. It is to eliminate recurring vulnerability classes, reduce exposure, and make the secure path the easiest path.

This role is based at our DNV office in Houston, TX or Oakland, CA, presenting a dynamic hybrid schedule where employees will typically spend three (3) days per week working from either a DNV office or client location/site. Further details regarding role-specific requirements will be shared during the interview process.

You'll be a technical leader within our organization focused on three core priorities:

• Securing application and AI architecture. Design and implement secure patterns across applications, APIs, cloud platforms, and AI-agent systems, with particular emphasis on identity, authorization, tenant isolation, data access, tool use, and runtime guardrails.
• Automating security in engineering workflows. Build and tune risk-based controls so material issues are caught and acted on inside delivery workflows, rather than at manual checkpoints.
• Eliminating recurring vulnerabilities. Find root causes, fix weaknesses at the architecture or platform-pattern level, and make the same class of issue structurally difficult to reintroduce

The responsibilities below describe how this work shows up day-to-day across architecture, delivery, AI systems, remediation, and engineering.

Build security into delivery and platform engineering

• Design and implement scalable controls for software and AI supply chains, including dependency integrity, SCA, SAST, DAST, build provenance, artifact security, secrets protection, container and infrastructure-as-code assurance, and software or AI bills of materials where appropriate.
• Implement platform-level controls: policy as code, authorization enforcement, data-access guardrails, secure defaults, and reusable reference implementations.
• Design AI-assisted security-testing environments, automated attack scenarios, and security-regression suites that prevent resolved issues from silently returning.
• Implement risk-based quality gates with documented exception paths, accountable ownership, and service-level expectations, so material issues block release.

Find, prove, and fix material weaknesses

• Review source code, APIs, and application designs for weaknesses in authentication, authorization, session management, input handling, data-access scope, and multi-tenant isolation, including row- and field-level boundaries.
• Conduct authorized application, API, and AI security testing, including targeted manual testing of business logic and trust boundaries that automated tools cannot adequately validate.
• Work alongside engineers to remediate root causes, validate fixes, create regression tests, and put preventive controls or secure patterns in place.
• Establish vulnerability triage and remediation practices, including exploitability and exposure analysis, accountable ownership, target dates, exception handling, retesting, closure evidence, and escalation of overdue material risk.

Secure AI agents and AI-assisted development

• Establish agent identities and least-privilege permissions, with clear separation of read, write, execute, approval, and administrative capabilities.
• Govern model, tool, skill, connector, plug-in, memory, and data access, including tenant isolation and boundaries between trusted and untrusted context.
• Validate untrusted inputs and tool outputs, and design defenses against direct and indirect prompt injection, goal manipulation, tool misuse, privilege escalation, sensitive-data exposure, memory poisoning, unsafe delegation, and cascading failures.
• Assess multi-agent workflows to implement approval requirements for consequential or irreversible actions, runtime policy enforcement, rate and resource limits, and tamper-resistant auditability.

Shape secure architecture at scale

• Lead high-risk threat modeling and architecture reviews for complex, multi-tenant, cloud-native, event-driven, and AI-enabled systems.
• Develop and demonstrate reusable secure patterns for microservices, APIs, event-driven systems, containers, Kubernetes, cloud services, and agentic AI applications.
• Contribute to…