Job Details

The AI Administrator plans, configures, secures, and monitors the platforms and controls that allow TRI to deliver AI-enabled and automation services safely and effectively. Working within TRI’s AI and Digital Enablement capability, the AI Administrator partners with the AI team (Architect, AI and Automation Engineers, and AI Business Liaisons), IT Infrastructure, Enterprise Application Development, and Information Security to ensure that Microsoft Foundry, Azure AI services, Microsoft 365 Copilot, Copilot Studio, Power Platform, and other approved AI services are provisioned in a governed, least-privilege, and auditable manner.  

The role owns day-to-day administration of AI platform access, privileged account authorizations, technical guardrails for acceptable use, data loss prevention, and AI usage and cost monitoring, translating TRI’s AI End User Policy and information security requirements into enforced, measurable technical controls.  

This is a platform administration and control role, not a solution development role. The AI Administrator is not expected to build AI applications or agents; AI and Automation Engineers build and evaluate solutions, and AI Business Liaisons qualify opportunities and support adoption. Instead, the Administrator provides the secure foundation those roles depend on: Foundry resource and project structure, role-based access and managed identities, model deployment and guardrail configuration, network and authentication controls, Microsoft Purview data protection, and monitoring through Azure Monitor, Application Insights, Microsoft Defender for Cloud, and Purview Audit.  

The Administrator ensures that AI capabilities move from prototype through pilot and production with appropriate access, data protection, logging, and cost visibility, and that TRI can produce evidence of these controls for client, ISO 27001, and federal (e.g., FISMA/NIST) security reviews.   

Experience administering Microsoft Azure, including subscriptions, resource groups, Azure RBAC, networking, Key Vault, and monitoring. 

Hands-on Microsoft Entra ID administration, including groups, app registrations, managed identities, Conditional Access, and MFA. 

Experience implementing least-privilege and privileged access controls, preferably with Entra Privileged Identity Management and access reviews. 

Working knowledge of Microsoft 365 security and compliance administration, including Microsoft Purview sensitivity labels, DLP, and audit. 

Experience configuring logging, monitoring, and alerting with Azure Monitor, Log Analytics, or equivalent tools, including basic KQL queries. 

Scripting and automation with PowerShell, Azure CLI, and/or Microsoft Graph. 

Understanding of generative AI concepts and risks (LLMs, agents, RAG, prompt injection, data leakage, oversharing) sufficient to configure and explain appropriate controls. 

Familiarity with security control frameworks (e.g., ISO 27001, NIST SP 800-53, NIST SP 800-171) and producing audit evidence. 

Strong documentation, communication, and cross-team coordination skills. 

Strongly Preferred 

Microsoft Foundry / Azure AI services administration, including Foundry RBAC, model deployments, and guardrails. 

Microsoft Purview DSPM for AI and DLP for Microsoft 365 Copilot and Copilot Studio. 

Microsoft 365 Copilot, Copilot Studio, and Power Platform administration and governance. 

Azure API Management / AI Gateway token limits, quotas, and usage metrics. 

Microsoft Defender for Cloud (including AI threat protection) and Microsoft Sentinel. 

Azure Policy and infrastructure-as-code (Bicep, ARM, or Terraform). 

Azure Cost Management, resource tagging, and chargeback/showback reporting. 

Microsoft certifications such as AZ-104, AZ-500, SC-300, SC-401, or AI-102. 

Federal contracting, clinical research, life sciences, or regulated/validated environment experience. 

Experience supporting FISMA, FedRAMP, CMMC, or ISO 27001 assessments and audits. 

Experience helping establish an AI governance, AI Center of Excellence, or AI platform operating model. 

Experience building SIEM detections or monitoring use cases specific to AI and agent activity. 

Demonstrated ability to tune guardrails and DLP controls to protect data without disrupting legitimate business and project work. 

This role is fully remote for those living in the DMV area. (Maryland, Virginia, Washington D.C.)