AI Integrity Engineer

Workforce Classification

Telecommuter

Our purpose, to improve customers’ lives by making healthcare work better, is far from ordinary. And so are our employees. Working at Premera means you have the opportunity to drive real change by transforming healthcare.

Premera is committed to be a workplace where people feel empowered to grow, innovate, and lead with purpose. By investing in our employees and fostering a culture of collaboration and continuous development, we’re able to better serve our customers. It’s this commitment that has earned us recognition as one of the best companies to work for.

Learn how Premera supports our members, customers and the communities that we serve through our Healthsource blog:  .

We are seeking an AI Integrity Engineer to strengthen security across our evolving AI, Cloud, and Data ecosystems. This role partners with Enterprise Engineering to architect and secure modern application and infrastructure ecosystems. By bridging the gap between Platform Engineering and AI Security, you will lead the development of secure AI pipelines, agentic workflows, and robust data warehouse protections. You’ll work cross‑functionally to build scalable, resilient foundations that embed AI‑specific guardrails directly into the enterprise fabric.

This is a hands‑on contributor role supporting teams building AI services by establishing secure identity, access, guardrails, and lifecycle controls for autonomous and semi‑autonomous AI agents.

This role may be hired at a Level III or IV, depending upon experience. This is a telecommuter position, working from home.

• SAST & DAST Implementation:
  Design and manage Static and Dynamic Application Security Testing pipelines to detect security flaws early in the lifecycle.
• Vulnerability Remediation:
  Lead regular vulnerability scans of codebases and containers. Analyze results, prioritize critical issues, and partner with engineering teams to drive remediation.
• Secure Supply Chain:
  Manage artifact security and dependency scanning using Artifact and Dependency Scanning tools.

• Protect AI Assets:
  Secure data pipelines, models, and agents from threats such as prompt injection, model hijacking, data‑poisoning, and trojaned instructions.
• Secure RAG Pipelines:
  Enforce retrieval integrity, document ingestion safety, and data‑access controls to prevent indirect prompt injection and data leakage.
• Secure Interaction:
  Enable secure agent‑tool interaction using frameworks such as MCP by implementing strong authentication, authorization, and scoped tool‑permission boundaries.
• Guardrails & Safety:
  Deploy and maintain AI guardrails, including safety filters, task adherence controls, scoped action permissions, and prompt‑shielding mechanisms.
• Threat Modeling:
  Perform AI‑specific threat modeling focused on jailbreaks, adversarial inputs, indirect prompt attacks, and non‑deterministic system behavior.
• Compliance:
  Support cross‑team compliance efforts by ensuring audit trails, usage governance, and adherence to emerging AI control frameworks.

• Implement identity governance frameworks tailored for AI agents, such as EPhemeral authentication, attribute‑based access control (ABAC), and just‑in‑time provisioning.
• Manage M2M/OIDC/OAuth identity configurations for per‑application agent authentication, including scopes, claims, client credentials, and automated key/secret rotation.
• Configure Azure Entra Agent IDs (or similar) to issue trackable, short‑lived identities for unique agents.
• Implement governance around credential lifecycle, preventing agent credential sprawl.

• Administer and maintain the MCP Registry or equivalent systems to govern secure access to tools and APIs for agent workflows.
• Monitor and govern agent tool selection, ensuring safe invocation boundaries and preventing privilege escalation within automated workflows.
• Support controlled integration of agents with internal services without owning traditional infrastructure security functions (e.g., firewall…