For the complete documentation index, see llms.txt. This page is also available as Markdown.

Suggestions

Search business concept suggestions (KNN)

post

Search for recommended business concept versions using KNN-based suggestion criteria.

Body

KNN vector search for business concept suggestions. Requires a resource (type + id) to generate embeddings. Optional parameters override similarity search behavior.

collection_namestringOptional

Embedding collection name (defaults to system default)

kinteger · min: 1 · max: 100Optional

Number of nearest neighbors to return (1-100)

Default: 10
num_candidatesinteger · min: 1Optional

Number of candidates for ANN search

Default: 100
similaritynumber · max: 1Optional

Cosine similarity threshold (0.0-1.0)

Default: 0.6
Responses
200

Suggestions list

application/json
post/api/business_concept_versions/suggestions
POST /api/business_concept_versions/suggestions HTTP/1.1
Content-Type: application/json
Accept: */*
Content-Length: 88

{
  "k": 10,
  "num_candidates": 100,
  "resource": {
    "id": 123,
    "type": "structures"
  },
  "similarity": 0.6
}
{
  "data": [
    {
      "business_concept_id": 55,
      "id": 101,
      "name": "Customer Segmentation",
      "similarity": 0.92,
      "updated_at": "2026-05-22T10:00:00Z"
    },
    {
      "business_concept_id": 56,
      "id": 102,
      "name": "User Clustering",
      "score": 0.88,
      "updated_at": "2026-05-22T09:50:00Z"
    }
  ]
}

Last updated