IT Security · Information Security · Enterprise AI Governance

What IT needs to know
before AI agents run
your operational workflows.

AI agents acting on operational workflows are not productivity tools. They initiate external system interactions, process sensitive data, make decisions on behalf of the enterprise, and generate audit events that compliance teams need to retrieve on demand.
The security review for an enterprise AI agent deployment covers the same surface areas as any production system review — plus the specific governance requirements of autonomous decision-making on regulated data.
This page answers the questions IT and information security teams ask in every enterprise AI vendor evaluation — not in general terms, but at the platform architecture level.
Every connection between your AI client and PLRX agents runs through MCP — an open standard with full logging, scoped permissions per tenant, and no proprietary dependencies. Every tool call is attributed to an identity and logged to the WORM audit trail. IT has complete visibility. Nothing runs outside the governed layer.

94% autonomous resolutionFrom $0.99 per missionEnterprise Agentic

Book a Scoping Call

Get the security answers.

Tell us which operational workflow is pending your security review. Proof of concept in 2–3 weeks — production in 12 weeks.

First Name Required.

Last Name Required.

Work Email Please enter your corporate email address.

Company Required.

Job Title Required.

Something went wrong. Please try again or email [email protected].

By submitting you agree to our Privacy Policy. We never sell your data.

Request received.

Someone from the PLRX team will be in touch within two business days to schedule your scoping call.

The IT Security Review — What Enterprise AI Agent Deployment Requires

These are the specific platform architecture questions IT and information security teams ask before approving an enterprise AI agent deployment. Each has a specific PLRX answer.

Access Controls and Authentication

Who can access what — and how is that enforced?

PLRX deployments use role-based access control (RBAC) with fine-grained permissions at the workflow, agent, and data level. SSO integration via SAML 2.0 and OIDC. Agent data access is scoped to the workflow at the infrastructure layer — the agent cannot access data outside its defined scope regardless of what it reasons its way toward. Service-to-service authentication via short-lived tokens with strict rotation. No long-lived credentials in the agent execution environment.

Data Residency and Tenant Isolation

Where does the data live — and can another customer's deployment touch it?

Sovereign per-tenant Kubernetes environment. Each customer's agents run in a dedicated environment with no shared runtime, no shared data plane, and no shared networking with other customer deployments. PHI and sensitive data are encrypted at rest with AES-256-GCM field-level encryption before persistence. All data in transit encrypted via TLS 1.3. Data residency contractually committed to AWS region. On-premises and VPC deployment available for organisations with strict residency requirements.

Audit Trail and Logging

What is logged — and who can retrieve it?

WORM (Write-Once Read-Many) append-only event log captures every agent action: what the agent read, what it decided, what it submitted, what it received, and when. Object-locked — cannot be modified or deleted. Every event attributed to a model, a timestamp, and an agent identity. Queryable by your security or compliance team without PLRX involvement. SIEM integration available via structured log export. Log retention configurable per compliance requirements.

Suspension and Incident Response

If something goes wrong, how fast can you stop it — and without calling the vendor?

Three-level suspension: platform-wide (halt all agents), agent-level (halt a specific agent type), workflow-level (halt a specific open mission). All three levels are immediate and do not require PLRX involvement. Suspended workflows preserve complete state for forensic audit. Incident response runbooks available for common scenarios. PLRX support SLA for security incidents: 1-hour response for critical, 4-hour for high.

Vulnerability and Penetration Testing

What is the security testing programme — and can we test it ourselves?

PLRX maintains SOC 2 Type II certification with annual third-party audits. Penetration testing conducted annually by independent security firms. Customer penetration testing permitted under coordinated disclosure programme with advance notification. Security findings disclosed to customers under responsible disclosure SLA. Dependency vulnerability scanning continuous via automated pipeline.

Security Architecture Reference — Platform Specifications

The security architecture
behind enterprise AI deployment.

Security Domain	PLRX Implementation	Applicable Standard / Requirement
Encryption at rest	AES-256-GCM field-level encryption on sensitive data before persistence. PHI encrypted at the database level before any logging occurs. Encryption keys managed per-tenant in AWS KMS.	HIPAA Technical Safeguard §164.312(a)(2)(iv); SOC 2 CC6.7
Encryption in transit	TLS 1.3 for all external communications. mTLS for service-to-service communication within the PLRX platform. Certificate rotation automated.	HIPAA Technical Safeguard §164.312(e)(2)(ii); SOC 2 CC6.7
Audit logging	WORM append-only event log. Object-locked. Every agent action logged with model attribution, timestamp, and action payload. Cannot be modified or deleted. Queryable without vendor involvement.	HIPAA §164.312(b); FINRA Rule 4511; SOC 2 CC7.2
Tenant isolation	Sovereign per-tenant Kubernetes namespace with dedicated compute, dedicated networking, and dedicated data stores. No shared runtime path between tenants. Isolation architectural — not logical partition.	SOC 2 CC6.6; ISO 27001 A.13.1.3
Agent authority enforcement	Workflow authority boundaries enforced at the PLRX platform layer — not in agent code. Agents cannot exceed defined data access scope or escalation thresholds at runtime. Platform enforcement is architecture-level.	NIST AI RMF GOVERN 1.1; SOC 2 CC6.3

IT Security · The Architecture Question That Determines Deployment Approval

What data does the agent touch — and who controls that boundary?

This is the IT security question that determines whether an enterprise AI agent deployment is approvable. The answer has two parts: what data the agent is authorised to access, and how that boundary is enforced.

PLRX answer — what the agent can access: Agent data access is scoped to the workflow at the infrastructure layer. A prior auth agent accesses clinical documentation relevant to the specific authorisation request. It cannot access the patient's full record, adjacent claims, or any data outside the defined workflow scope — because the platform does not grant that access.

PLRX answer — how the boundary is enforced: Platform architecture, not agent code. The access boundary is defined in the workflow configuration and enforced at the infrastructure layer before the agent can request data outside its scope. It is not a guideline the agent is expected to follow — it is a hard boundary the platform enforces.

Tenant isolation means data from one customer deployment cannot be accessed by another. Customer data never enters model training pipelines — contractual. The complete security architecture documentation, SOC 2 report, and penetration testing summary are available for security review teams on request.