v18

latestOpenAPI 3.0.0raw.githubusercontent.com2024-11-18167108845.6 KB

Query using RAG Search

Query across your connected data sources using RAG Search

post/rag/query

Headers

x-connection-tokenstring required

The connection token

Request body

querystring required

The query you want to received embeddings and chunks for

topKnumber nullable

The number of most appropriate documents for your query.

Example request

{
  "query": "When does Panora incorporated?"
}

Response

chunkstring required

The chunk which matches the embed query

metadataobject nullable required

The metadata tied to the chunk

scorenumber nullable required

The score

embeddingnumber[] nullable required

The embedding of the relevant chunk

Example response

[
  {
    "chunk": "Date : 06/07/2023",
    "metadata": {
      "blobType": "",
      "text": "ATTESTATION"
    },
    "score": 0.87,
    "embedding": [
      -0.00442447886,
      -0.00116857514,
      0.00869117491,
      -0.0361584462,
      -0.00220073434,
      0.00946036354,
      -0.0101112155
    ]
  }
]