Documentation

Platform overview

AiVRIC is an autonomous security, compliance, and risk intelligence platform that runs in your environment. This page covers the architecture, modules, guardrails, and operational model.

Architecture

AiVRIC is composed of four interconnected modules deployed as containerized services on your Kubernetes cluster. All processing stays in your environment — no data leaves your control plane.

CloudSignals+RiskOps™

Continuous cloud posture monitoring, compliance automation, and AI-driven risk remediation across AWS, Azure, GCP, Kubernetes, and more.

AI Signals™

Real-time detection of AI model misuse, shadow AI, prompt injection, PII leakage, and model drift across your AI workloads.

AIRE Agentic Mesh™

Autonomous remediation engine that closes security gaps, raises tickets, and applies approved fixes without human handoff.

RogueAgent ASPM™

Application Security Posture Management for agentic and AI-driven applications — tracks permissions, dependencies, and runtime behavior.

Vision AI Optics™

AI/ML-powered threat intelligence and behavioral analytics that surface anomalies before they become incidents.

Shared Data Layer

A unified data fabric that normalizes signals from every connected source into a single risk-aware context model.

AiVRIC CloudSignals+RiskOps Command Center showing posture threat score, findings counts, threat map with global signal coverage, and control plane activity — **CloudSignals+RiskOps in operation** — the module's command center shows the live threat score, active findings across all connected providers, threat map, and posture control plane panels updated every 24-hour scan cycle.

Connectors

AiVRIC connects to your environment via read-only, least-privilege integrations. No agents required for cloud accounts.

Category	Supported providers
Cloud providers	AWS, Microsoft Azure, Google Cloud Platform, Alibaba Cloud
Container & Kubernetes	Amazon EKS, Azure AKS, Google GKE, self-managed clusters
Source control	GitHub, GitLab (roadmap)
SaaS productivity	Microsoft 365 / Entra ID
Databases	MongoDB Atlas
Ticketing & alerting	Jira, ServiceNow, PagerDuty, Slack, Microsoft Teams
SIEM	Splunk, Microsoft Sentinel, Elastic SIEM (via webhook)

Guardrails

Guardrails are policy-as-code rules that AiVRIC evaluates continuously. They can run in Detect (alert only), Prevent (block + alert), or Remediate (auto-fix + alert) mode.

Always start in Detect mode and review impact for at least one scan cycle before promoting to Prevent or Remediate.

Compliance packs

Pre-built control mappings for SOC 2 Type II, ISO 27001, PCI DSS, HIPAA, NIST CSF, CIS Benchmarks, CMMC, and more.

AI safeguards

Detect PII exfiltration, model hallucinations, prompt injection, unauthorized model access, and jailbreak attempts in real time.

Custom rules

Author guardrails in YAML or Python. AiVRIC evaluates them within the same scan cadence as built-in rules.

Deployment model

AiVRIC deploys into your Kubernetes cluster using Helm charts managed by ArgoCD. All images are pulled from your private container registry — no public internet egress required at runtime.

AiVRIC requires cluster-admin permissions during initial Helm install only. Runtime permissions are scoped to the aivric namespace via RBAC.

Asset inventory dashboard showing 319 discovered resources across connected cloud providers with ownership and sync status — **Asset inventory** — every cloud resource discovered across connected providers, with ownership, criticality, and sync freshness tracked automatically.

Findings list showing all detected control failures with severity, provider, resource, and status columns — **Findings** — all detected control failures classified by severity, tagged to provider and resource, and available for immediate triage and treatment.

Data handling

AiVRIC processes configuration metadata — not customer data. All findings, scan results, and audit logs are stored in your environment's PostgreSQL instance and are never transmitted to AiVRIC infrastructure.

Data residency

100% in-environment. Your cloud, your cluster, your data store. AiVRIC never receives your findings or cloud configurations.

Encryption

TLS 1.3 in transit; AES-256 at rest via your cloud provider's KMS. Secrets stored in Azure Key Vault / AWS Secrets Manager.

Audit logging

Every platform action is logged with actor, timestamp, and outcome. Logs forward to your SIEM via configurable webhook.