Brand Control Plane
Real-time brand conformance scoring API — score products, outfits, and assortments against brand DNA
The Brand Control Plane is the operational API layer of the platform — where brand DNA moves from analysis into action. It provides a real-time scoring engine that evaluates products, outfits, and assortments against brand DNA, returning congruence decisions that can be integrated into buying workflows, PIM systems, and e-commerce platforms.
The scoring engine uses a four-layer architecture: (1) Rule pre-filter checks products against categorical outfit and context rules for instant pass/fail on hard constraints. (2) Statistical scoring computes cosine similarity and Jensen-Shannon divergence between product attributes and brand distributions. (3) Policy evaluation applies configurable thresholds to classify products as CONGRUENT, WARNING, or INCONGRUENT. (4) Ensemble combines all layers into a final score with configurable weights.
The live scoring console lets you test the API interactively. Enter product attributes, select a brand, and click Score to see the full scoring breakdown — overall score, decision, per-attribute match details, and any rule violations. This is useful for understanding how the scoring engine evaluates specific products before integrating the API into your systems.
API health monitoring shows the current status of the scoring engine, database connection, and rule engine. Uptime, response latency, and active API key count are tracked in real time. The health endpoint at /api/v1/health is the standard integration health check.
Authentication uses API keys with tiered rate limiting. Explorer tier (10 req/min) is for testing and development. Professional tier (100 req/min) is for production integration. Enterprise tier (1000 req/min) is for high-volume real-time scoring. Keys are managed through the admin API and hashed with SHA-256 for security.
The API reference section documents all available endpoints with request/response formats. The scoring endpoints accept product attributes as JSON and return structured decisions that downstream systems can act on — for example, flagging a new product in the PIM as 'WARNING: low brand alignment' before it goes to production.
Tips
- Use the live scoring console to test edge cases before going live — products at the boundary of CONGRUENT/WARNING thresholds are the most important to validate.
- The X-Scoring-Duration-Ms header on every response lets you monitor latency. Single product scoring should be under 30ms.
- Start with Explorer tier for development, upgrade to Professional for staging/UAT, and Enterprise for production traffic.
- The scoring weights (statistical: 0.4, rule: 0.35, context: 0.25) can be customised per brand to reflect different governance priorities.
- Integrate the /api/v1/health endpoint into your monitoring system for real-time API status.
- Review the scoring audit log periodically to understand how the engine is classifying products and whether thresholds need adjustment.
Live Scoring Console
Score a product against brand DNA in real-time
Brand DNA Profile
API Reference
/api/v1/scoreScore product(s) against brand DNA/api/v1/score/outfitScore outfit combination compatibility/api/v1/score/assortmentBatch score collection (up to 500 items)/api/v1/brand/{id}/profileGet brand distribution profile/api/v1/brand/{id}/rulesList active rules for brand/api/v1/brandsList all brands/api/v1/healthHealth check with uptime and statusAuthentication
Include X-API-Key: your_key or Authorization: Bearer your_key header. Three tiers: Explorer (10/min), Professional (100/min), Enterprise (1000/min).