Compliance Command Center

NIST AI RMF Governance Dashboard

Real-time trust evidence layer mapping Sentinel Docs controls to Govern, Map, Measure, and Manage requirements.

Compliance Health

No audit evidence yet: run Q&A in Secure RAG Shell to generate baseline scores.

BASELINING

Sample Size

Latest Judge Score

Average (Last 10)

Static NIST AI RMF alignment map for Sentinel controls.

NIST AI RMF compliance mapping of Sentinel controls, evidence, and audit details.
NIST Function	Requirement	Evidence	Audit Detail
GOVERN (GV)	Accountability & Transparency	LangSmith Traces	Immutable audit trail of every interaction.
MAP (MP)	Contextualizing AI Use Cases	Upstash Namespaces	Isolated session-level document context.
MEASURE (MS)	Assessing System Trustworthiness	LLM-as-a-Judge	Real-time faithfulness and grounding scores.
MANAGE (MG)	Prioritizing & Acting on Risks	Upstash Redis + Sentinel Guard DLP	Edge rate limiting and redaction trigger controls.

Total PII patterns blocked in the active session.

Open session trace evidence and export compliance report artifacts.

Session: No active session

Visual summary of the last 10 Judge scores from this browser session.

No judge history yet. Ask a few questions in the chat to populate this heatmap.

Compliant (>= 0.90)Review (0.70 - 0.89)Violation (< 0.70)

Session token and cost telemetry (UI scaffold; LangSmith wiring follows).

GPT-4o-mini

Total Tokens Consumed

In: 0 / Out: 0

Estimated Session Cost

$0.0000

Pricing: $0.15/M input - $0.60/M output

Cost-to-Intelligence Efficiency

Awaiting judged response to compute efficiency

Side-by-side model KPI scaffold (LangSmith eval wiring to follow).

Benchmark comparison of model faithfulness, latency, and cost per response.
Model	Faithfulness	Avg. Latency	Cost / Response	Status
GPT-4o-mini	0.98WINNER	0.80sWINNER	$0.0020WINNER	active
Llama 3.3 70B	--	--	--	pending