What it does: Assigns each supplier room to one reference bucket.

Workflow role: Alternate sync view (not step 0 of the main flow).

Use when: You want reference-centric grouping instead of per-supplier scores.

Do not use when: You need ranked candidates or async persistence.

Details, examples, limits, and result codes: Implementation Workflow (doc page: implementation-workflow).

{
  "hotelId": "hotel-1",
  "chunkKey": "chunk-1",
  "threshold": 0.8,
  "references": [
    {
      "name": "King Room"
    }
  ],
  "suppliers": [
    {
      "name": "Deluxe King"
    }
  ]
}

{
  "data": {
    "hotelId": "hotel-1",
    "supplier_assignment": "best_reference_exclusive"
  }
}