The company develops clean, high-temperature nuclear power designed to scale. Its first pilot plant is in Orangeville, Utah.
You would join IT & Enterprise Software, part of the Business organization, which runs the technology infrastructure, enterprise applications and digital tools behind the company's operations.
Responsibilities
Build and run application and AI security for the company's internal platforms, AI agents, generative AI models and knowledge bases, tools, connectors, plug-ins, APIs, code and repositories, hosted in AWS and Microsoft Azure (commercial and GCC High).
Put security controls and tooling in place across the software lifecycle, including SAST, SCA, DAST, IaC scanning and AI-specific scanning and red teaming. Software Engineering fixes the findings.
Work with developers so secure design is the normal approach, keep controlled data within authorized boundaries, and give leaders clear measures of risk.
Implement the NIST Secure Software Development Framework through coding standards, review gates, vulnerability response timelines and developer training.
Deploy, tune and run code, dependency and posture scanning, such as Checkmarx One and Gov, across every repository, sending controlled data only to FedRAMP services.
Run authenticated dynamic testing and schema-based API testing on web and API surfaces, before release and on a schedule.
Secure workloads, containers, serverless functions and AI services in AWS and Azure, scan Terraform, Bicep and Kubernetes definitions, and act on posture findings.
Threat-model and secure generative AI systems against the OWASP Top 10 for LLM Applications, covering prompt injection, insecure output handling and excessive agency.
Define permission models for AI agents: scoped tool access, human approval for consequential actions and kill switches. Vet connectors, plug-ins and MCP servers.
Govern the Microsoft Entra enterprise applications and connectors AI tools use to reach company data, covering OAuth consent, least-privilege permissions and access reviews.
Build adversarial test suites and evaluation gates for agents and models, and verify model and dataset provenance before release.
Make sure retrieval and model access respect data classification and source permissions. Keep CUI and export-controlled data in authorized services and the correct tenant, commercial or GCC High.
Enforce secrets management and rotation, secret scanning, dependency pinning, SBOM generation and artifact signing.
Maintain one prioritized backlog across all scan types, with severity, owner and SLA. Triage it weekly with Software Engineering and report open risk and time to remediate.
Define logging for agent actions, tool calls and model calls, forward it to the SIEM/XDR platform, and build detections for misuse.
Extend secure configuration and scanning to on-premises engineering systems, build infrastructure and developer workstations.
Apply the NIST AI RMF, NIST SP 800-171 / CMMC Level 2 and DOE O 205.1D to AI and software systems. Maintain the AI tool inventory and acceptable-use policy.
Requirements
4+ years in application, product or cloud security engineering. The senior level requires 7+ years.
Hands-on experience securing LLM-based or agentic AI applications, including threat modeling.
Experience running SAST, SCA and DAST tools, such as Checkmarx or similar, inside CI/CD pipelines.
Python plus at least one other language used in modern services, such as TypeScript, Go or Java, and the ability to read and review code.
Strong security knowledge of AWS and Azure covering identity, key management, networking and logging.
Experience with Terraform, IaC scanning and container security basics.
Working knowledge of the OWASP Top 10, the OWASP Top 10 for LLM Applications and secure API design with OAuth 2.0 / OIDC.
Experience with secrets management, dependency management and supply chain controls such as SBOMs and signing.
Ability to turn findings into prioritized guidance developers can act on, communicating clearly in writing and out loud.
You must be a U.S. person under U.S. export control rules.
Preferred: startup or high-growth company experience, and comfort working at a fast pace.
Preferred: self-starter who can take on projects and deliver with little oversight.
Preferred: hands-on work with DOE O 205.1D, CMMC and NIST CSF 2.0.