Platform Architecture

Four module groups.
One integrated control plane.

Equanimo protects enterprise knowledge by understanding the full context behind every AI interaction—not just prompts, policies, or keywords.

It evaluates user identity, personas, data sensitivity, business context, and organizational policies in real time to prevent unauthorized knowledge exposure. Built on a plug-and-play architecture, Equanimo integrates seamlessly with existing AI applications, APIs, and models without requiring application changes.

Request Demo Explore Use Cases
Platform Overview

The full platform, at a glance.

The live Platform page content is preserved here in a static, readable layout so the page renders reliably while keeping the same Home page typography, color language, and CTA styling.

Layer 04Command

Organizational Sovereignty

Real-time visibility and control across your AI estate.

CISOCROCCO
Layer 03Evidence

Risk Artifacts

Quantified, auditable risk intelligence from every interaction.

CISOVP InfoSecGeneral Counsel
Layer 02Policy

Governance

Define rules, enforce them, prove they were followed.

CCOGeneral CounselIT RiskCISO
Layer 01Foundation

Data Collection

The layer every other module depends on.

Security EngITDevOps
Layer 04 Command

Organizational Sovereignty

One view answers the question every CISO starts their day with: are we safe right now? Real-time visibility and instant action across your entire AI estate.

Primary users
CISOCROCCO
1.1 · Primary entry point

Dashboard

Compliance scores, risk flags, AI adoption and model activity — consolidated across every team into a single real-time view.

  • Composite compliance score across GDPR, HIPAA, SOC 2, ISO 42001 and EU AI Act — with 7d / 30d / 90d trend
  • AI adoption heatmap: approved vs shadow AI by team, live from Gateway traffic
  • Top policy violations ranked by 24h trigger count with team attribution
  • Cost & usage: prompts inspected, tokens processed, estimated AI spend
1.2 · Operational console

Control Tower

Every AI interaction as it happens — with the power to act immediately: suspend access, invoke a policy, or escalate an incident.

  • Live prompt stream: every request transiting the Gateway in real time
  • Inline policy execution — block, warn, quarantine or redirect without leaving the feed
  • One-click user or team suspension, effective immediately at the Gateway layer
  • Incident escalation with reviewer assignment, SLA tracking, and full audit trail
Layer 03 Evidence

Risk Artifacts

Quantified, auditable risk intelligence built from every AI interaction — detecting exposures in content, not just metadata.

Primary users
CISOVP InfoSecGeneral Counsel
3.1 · Forensic log

Prompt Audit

Complete, searchable record of every AI interaction — content-level detection of sensitive data, not just metadata.

  • Full prompt log with PII, PHI, financial and source-code detection before storage
  • Policy violation log with matched rule, triggering fragment and outcome
  • Searchable by user, team, model, entity type and date
  • Role-gated raw prompt access — with access itself written to the audit trail
3.2 · Quantified exposure

Risk Scoring

Continuous, evidence-based risk scores for every model, team and use case — a live signal that drives remediation.

  • Multi-factor composite score (0–100) with automatic escalation at Critical tier
  • Remediation priority queue with suggested actions and drill-through
  • Score history — prove governance actions are reducing exposure
3.3 · IP protection

Knowledge Shield

Real-time detection and blocking of proprietary information — source code, pricing models, client data — in prompts and responses.

  • Knowledge fingerprinting — custom sensitive categories detected without exposing the content
  • Bidirectional detection on outbound prompts and inbound AI responses
  • Vector DB analysis — scheduled and on-demand scans of knowledge bases
Layer 02 Policy

Governance

Define the rules that govern AI use, enforce them automatically, and generate the evidence that proves they were followed — continuously, not just at audit time.

Primary users
CCOGeneral CounselIT RiskCISO
2.1 · Source of truth

Policy Centre

Author, version and publish AI governance rules as machine-readable controls — active at the Gateway the moment they're approved.

  • Structured authoring with full version control, approval workflows and immutable change history
  • Scope targeting at org, BU, team or model level with inheritance rules
  • Simulate a policy against historical prompt data before it goes live
  • Controls linked to the regulatory articles they satisfy
2.2 · Enforcement

AI Guardrails

Real-time enforcement on every prompt and response — blocking, redacting or flagging before data moves.

  • Pre-built library covering PII, source-code leakage, financial data, credentials and prohibited use
  • Custom rule builder with pattern matching and ML classifiers
  • Bidirectional enforcement on both inbound prompts and outbound responses
  • Per-rule action: hard block, silent redaction, warn-and-allow or log-only
2.3 · Continuous scoring

Compliance Adherence

Continuous, automated measurement against six regulatory frameworks — with one-click evidence packages ready for auditors.

  • Live coverage dashboard for GDPR, HIPAA, SOC 2, ISO 42001, EU AI Act and NIST AI RMF
  • Control-to-policy mapping with gap identification and remediation guidance
  • One-click audit evidence package — formatted for regulatory submission
2.4 · External risk

Vendor Compliance

Governance of every external AI vendor — scored against your framework, with automatic alerts when their terms change.

  • Vendor registry auto-populated from Gateway traffic, with compliance scorecards per vendor
  • DPA and contract tracker with permitted data categories, residency and expiry monitoring
  • Alerts when terms, privacy policy or certifications change
Layer 01 Foundation

Data Collection

Gateway intercepts and governs AI traffic at the network. Connectors extend that visibility to tools outside the network boundary. The layer everything else depends on.

Primary users
Security EngITDevOps
4.1 · Enforcement point

AI Gateway

A policy-enforced proxy between your organisation and every AI provider — every API call classified, governed, logged, and routed in real time.

  • Unified proxy for OpenAI, Anthropic, Azure OpenAI, Gemini, Bedrock and internal models — no SDK changes required
  • Authentication via your existing IdP over OIDC / SAML
  • Real-time governance pipeline: classification, PII detection and guardrail evaluation at every request
  • Rate limits, spend quotas, model routing, failover and data-residency region lock
4.2 · Integrations

Connectors

Extends Equanimo's visibility to AI tools that can't be proxied — Microsoft Copilot, GitHub Copilot, Salesforce Einstein.

  • Pre-built connectors for Microsoft Copilot for M365, GitHub Copilot and Salesforce Einstein
  • SIEM integration with Splunk, Microsoft Sentinel and Datadog
  • Identity provider sync — keep user rosters and roles consistent across the platform
Platform Architecture

How Equanimo fits in.

Equanimo protects enterprise knowledge by understanding the full context behind every AI interaction—not just prompts, policies, or keywords. Built on a plug-and-play architecture, it integrates with existing AI applications, APIs, and models without requiring application changes.

Equanimo platform architecture diagram
Equanimo platform architecture diagram
Get Started

See every module in a working environment.

A 45-minute demo against your own risk profile — focused on the modules most relevant to your compliance posture.

Request Demo See the Rollout Plan