AI Security Architect
Job Description:
Identity: NHI & AI securityReference architecture for AI agent identity: registration, attestation, credential issuance, scoping, revocationDelegation and on-behalf-of patterns — token exchange, audience restriction, downscoping across multi-agent call chainsTool/function invocation authorization: per-tool entitlement, policy enforcement at the agent or LLM gateway, approval gates for high-consequence actionsMCP and tool-protocol security — server trust, tool-definition integrity, confused-deputy preventionContainment patterns that assume the reasoning layer is compromised and enforce least privilege at the identity layerWorkload identity to replace long-lived secrets: SPIFFE/SPIRE, cloud workload identity federation, mTLS, short-lived credentialsNHI discovery, ownership, recertification, and deprovisioning integrated into existing IGABehavioral telemetry for agent and workload identity misuse; threat models via MITRE ATLAS and OWASP LLM Top 10 Enterprise architectureSecurity reference architecture, patterns, standards, exception governanceZero trust across identity, device, network, app, dataIAM/IGA/PAM, multi-cloud security, DevSecOps and supply chain, cryptography and key managementExecutive and audit advisory; mentorship of senior architectsRequired 12+ years security, 5+ in senior/principal architecture at enterprise scaleDeep identity architecture ownership across workforce, privileged, and machine identityCommand of OAuth 2.0/2.1, OIDC, JWT, RFC 8693 token exchange, mTLS, ABAC/ReBAC, policy-as-codeHands-on securing production AI or agentic systemsSecrets management and workload identity platforms; cloud security depth in two or more providersCodes well enough to prototype and review application code in design sessionsPreferredProduction MCP servers, agent frameworks, or LLM gatewaysIETF WIMSE and emerging machine identity standardsNHI security tooling; machine identity in SailPoint/SaviyntRegulated industry experience (FedRAMP, DFARS/CMMC, PCI DSS, NERC CIP)