Decision Orchestration Platform

Governance layer for
AI-driven decisions

Kirtonic sits between your analytics and action. Apply rules, enforce human approval, and maintain complete audit trails — without changing your existing pipelines.

Get Started How It Works

The Decision Layer

Your ML models produce predictions. Kirtonic decides what to do with them.

Data Source

BigQuery, Vertex AI, Databricks, or any API

2
Kirtonic
Rules engine, approvals, audit trail

Action

Webhooks, notifications, workflows

Why Kirtonic

AI models are powerful. But who decides what to do with their outputs?

Human-in-the-Loop

High-stakes decisions require human approval. Low confidence? High severity? A human reviews before action.

Configurable Rules

Define severity thresholds, approval requirements, and actions per workspace. No code changes needed.

Complete Audit Trail

Every signal, decision, approval, and action is logged. Full traceability for compliance and debugging.

Zero Data Movement

Kirtonic reads outputs, not raw data. Your sensitive data stays where it is.

Simple Integration

Send a signal, get a decision. One API call.

POST /v1/signals

curl -X POST https://api.kirtonic.io/v1/signals \
  -H "X-API-Key: sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "entity_id": "STORE_001|SKU_042",
    "risk_score": 0.85,
    "confidence": 0.72
  }'

# Response
{
  "status": "awaiting_approval",
  "severity": "high",
  "reason": "confidence below threshold"
}