Pre-dispatch validation.
Every AI utterance passes through one gate before it reaches the user. Brand governance, commercial validation, scope-of-practice enforcement, PII compartmentalisation. Target latency under 10ms via Go-native hot-path.
POST /v1/governance/gate · brand-governance example
curl -X POST https://api.audact.ai/v1/governance/gate \
-H "Authorization: Bearer $AUDACT_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"agent_id": "agent_dental_intake_v2",
"candidate_response": "Sure, I can offer you 50% off the implant package.",
"policy_bundle": "dental_de_brand_v3",
"context": {
"caller_role": "patient",
"session_id": "sess_8h2k9d7"
}
}'200 OK · verdict + rewrite suggestion
{
"verdict": "block",
"reasons": [
{
"policy": "commercial_commitment",
"rule": "no_discount_without_authorization",
"severity": "hard"
},
{
"policy": "brand_governance",
"rule": "approved_pricing_only",
"severity": "hard"
}
],
"rewrite_suggestion": "I can connect you with our scheduling team to discuss available packages.",
"latency_ms": 7.2,
"receipt_id": "rcpt_gov_01HX7..."
}Approved messaging, tone, claims. No off-script discounts, no invented product features.
Price, term, and commitment validation. Block or escalate before the user hears a number.
Medical, legal, financial advice gating. Route to a qualified human when the model strays.
Crypto-shredding, per-session keys, redaction. No PII leaves its consent envelope.
Included in Agency Pro & Enterprise — book a call and we'll scope it with you.