Skip to main content

Dashboard

The Beak AI Security Dashboard provides a role-scoped overview of your AI security posture. Four personas are available — each surfacing the metrics, widgets, and quick actions most relevant to that role. You can switch between views using the tabs at the top of the page.

AI/ML Engineer​

The AI/ML Engineer view focuses on the technical health of registered agents and the security testing pipeline.

Top metrics:

  • Registered Agents — Total agents in the registry with active/suspended breakdown and health status.
  • AIBOM Components — Number of AI systems tracked in the AI Bill of Materials.
  • Red Team Campaigns — Total campaigns run, with completed count highlighted.
  • Active Validators — Total configured validators across all policies.

Widgets:

  • Agent Registry — Quick-view list of agents with their risk scores and active/suspended status. Links to the full Agent Registry.
  • Recent Campaigns — Latest red team campaigns with their status (Pending, Completed). Links to Red Teaming.
  • Forensic Sessions — Most recent interaction replay sessions. Highlights recent violations inline. Links to the Forensic Explorer.
  • Technical Security Graph — Bar chart summarizing Agents, Validators, Campaigns, AIBOMs, ATLAS coverage percentage, average risk score, scan-ready systems, and session count.

CIO / Executive​

The CIO / Executive view provides a high-level security posture summary for leadership.

Top metrics:

  • Security Posture — An aggregate score (e.g., 91% — Strong) combining policy coverage, ATLAS coverage, validator health, and chain integrity.
  • ATLAS Coverage — Percentage of MITRE ATLAS techniques covered by the platform's defenses.
  • Registered Agents — Total agent count.
  • Total Events — Cumulative moderation events recorded in the ledger.

Widgets:

  • Policy Coverage — Count of active policies, guardrails, and validators.
  • Red Team Status — Breakdown of total, completed, and currently running campaigns.
  • Supply Chain & Integrity — AIBOM count and ledger integrity verification status.
  • Threat Mitigation Timeline — Stacked bar chart showing daily moderation outcomes over time.
  • System Integrity Score — Composite score breaking down Chain Integrity, Policy Coverage, ATLAS Coverage, and Validator Health individually.

Compliance Officer​

The Compliance Officer view centers on regulatory posture, review backlog, and ledger integrity.

Top metrics:

  • Compliance Rate — Overall percentage of compliant events. A Critical badge appears when the rate falls below threshold.
  • Policies Active — Count of active policies with total guardrails linked.
  • Pending Reviews — Number of interactions currently in the Review Queue awaiting human decision.
  • Audit Trail — Total events in the tamper-proof ledger with chain verification status.

Widgets:

  • Ledger Integrity — Real-time chain verification showing how many events have been verified against the hash chain.
  • Moderation Breakdown — Donut breakdown of Compliant, Violation, and Pending Review event counts.
  • Policy Enforcement Map — Full list of all active policies, each showing its guardrail count and enforcement status.
  • AIBOM Inventory — List of AI systems tracked in the supply chain, with versions.

SecOps Engineer​

The SecOps Engineer view is a live operational feed for the security operations team.

Top metrics:

  • Total Validators — Number of configured validators.
  • Active Guardrails — Number of active guardrails.
  • Policies Enforced — Count of policies currently active.
  • Risk Score Avg — Average risk score across all moderation events (0–100 scale).
  • Review Queue — Current count of pending human reviews.
  • Moderation Events — Total events processed.

Widgets:

  • Recent Moderation Events — Live feed of the latest AI interactions showing the input prompt snippet, Allow or Review outcome, risk score, and timestamp.
  • Ledger Integrity — Chain verification status with event count.
  • Quick Actions — Direct links to the Review Queue, Audit Log, and Validator management.

Best Practices​

  • Use the CIO / Executive view for board-level or management reporting — the System Integrity Score provides a single exportable metric.
  • Set a routine to check the Compliance Officer view for pending reviews — unreviewed items in the queue can affect your compliance rate.
  • Keep the SecOps Engineer view open during active deployments to catch policy violations in real time.
  • Use the AI/ML Engineer view when debugging agent behavior — the Forensic Sessions widget links directly to step-by-step replays of flagged interactions.