Apply on Employer Site

Datavant · 8 hours ago

Staff AI Security Architect

United States

Full-time

Remote

Lead/Staff

$224K/yr - $280K/yr

8+ years exp

Datavant is a data platform company and the world’s leader in health data exchange. As a Staff AI Security Architect, you will lead the design and adoption of secure principles across AI and Machine Learning systems, ensuring that these systems are secure, trustworthy, and scalable.

BiopharmaClinical TrialsData IntegrationHealth CareSoftware

No H1B

Responsibilities

Establish AI security architectural standards, design patterns, and best practices adopted across engineering teams

Architect and advise on secure end-to-end AI systems, including data pipelines, model training, evaluation, deployment, runtime monitoring, and agentic workflows

Lead threat modeling, architecture reviews, and risk assessments for AI-driven products, including LLMs, agent frameworks, and multi-agent systems

Define and evolve a comprehensive AI/ML secure development lifecycle integrated into existing SDLC practices

Develop reference architectures, documentation, and reusable security components to accelerate secure AI adoption

Collaborate with legal, privacy, compliance, and responsible AI stakeholders to align security controls with regulatory and ethical requirements

Act as a trusted advisor to senior leadership on AI security risks, trade-offs, and long-term strategy

Qualification

AI security architectureAI/ML systems securityThreat modelingRisk assessmentsSecurity controlsHealthcare security experienceAgentic AI workflowsAI threat modelsLarge-scale systems designCommunication skills

Required

8+ years of experience in security architecture, application security, or product security, with meaningful focus on AI/ML systems

Hands-on experience securing AI/ML or LLM-based systems, including familiarity with modern AI architectures and agentic workflows

Strong understanding of AI threat models, including adversarial ML, prompt injection, data poisoning, model theft, and abuse scenarios

Proven ability to design and influence security architectures for large-scale, distributed systems

Strong communication skills with the ability to translate complex technical concepts to both technical and non-technical stakeholders

Have a strong understanding of security controls, both those that exist in audit standards as well as practical controls that can help reduce risk and increase safety

Preferred

Hands-on experience building, testing, or tinkering with agentic AI workflows, with an understanding of the security risks they introduce

Experience securing AI/ML workloads in Databricks, with a deep understanding of its platform-specific security risks

You have experience with security in healthcare or other highly regulated space. Examples: HIPAA, HITRUST, SOC 2, PCI, FedRamp experience from an operational response standpoint