POST
/
api
/
v1
/
dataset
/
upload
/
platform
Upload New Dataset Platform
curl --request POST \
  --url https://api.collinear.ai/api/v1/dataset/upload/platform \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: multipart/form-data' \
  --form 'dataset_name=<string>' \
  --form space_id=3c90c3cc-0d44-4b50-8888-8dd25736052a \
  --form 'evaluation_type=<string>' \
  --form skip_context_check=false \
  --form file=@example-file
{
  "message": "<string>",
  "data": {}
}

Overview

Use this endpoint when you want rows to skip the regular dataset list and go straight into the Assess dashboard. It is the first call in the Assess Conversations workflow.

Form fields

  • file (required): JSON file containing an array of rows with conv_prefix and response objects. Use application/json for the MIME type.
  • dataset_name (required): Friendly name that appears in the dashboard.
  • space_id (required): UUID for the destination space.
  • evaluation_type (required): Typically safety; determines which judge templates Collinear recommends.
  • skip_context_check (optional): Set to true to skip automatic field validation while prototyping.

Example request

curl https://stage.collinear.ai/api/v1/dataset/upload/platform \
  -H 'Authorization: Bearer <token>' \
  -F 'file=@conversations.json;type=application/json' \
  -F 'dataset_name=assess-demo' \
  -F 'space_id=<SPACE_ID>' \
  -F 'evaluation_type=safety' \
  -F 'skip_context_check=true'

Response

On success the API returns a JSON body similar to:
{
  "message": "Dataset uploaded successfully",
  "data": {
    "dataset_id": "<DATASET_ID>",
    "eligible_evaluation_types": {"evaluation_types": ["safety", "performance"]},
    "rows": 2
  }
}
Persist the dataset_id for the follow-up judge creation and assess run calls.

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

multipart/form-data
file
file
required
dataset_name
string
required
space_id
string<uuid>
required
evaluation_type
string | null
skip_context_check
boolean
default:false

Response

Successful Response

message
string
required
data
object | null