Sign in

v5

latestOpenAPI 3.1.02026-08-025631,1012.8 MB
Scaffold Templates

Instantiate Scaffold Template

Create complete infrastructure from scaffold template.

Creates all resources atomically:

  1. Namespace with configured feature extractors
  2. Bucket with schema for your data structure
  3. Collection linked to bucket with feature config
  4. Retriever with search pipeline stages

All resources are empty, ready for data upload.

Next Steps:

  1. Upload data: POST /v1/buckets/{bucket_id}/objects
  2. Process batch: POST /v1/collections/{collection_id}/batches
  3. Search: POST /v1/retrievers/{retriever_id}/retrieve

Example Request:

{
    "namespace_name": "my_video_app",
    "namespace_description": "Video search application"
}
post/v1/templates/scaffolds/{template_id}/instantiate

Path parameters

template_idstring required

Scaffold template ID

Scaffold template ID

Request body

namespace_namestring required

Name for the new namespace (required, must be unique)

namespace_descriptionstring nullable

Optional description for the namespace

bucket_namestring nullable

Override default bucket name from scaffold

collection_namestring nullable

Override default collection name from scaffold

retriever_namestring nullable

Override default retriever name from scaffold

include_sample_databoolean

If true, include sample data from demo namespace. If false, create empty resources.

Example request

{
  "namespace_name": "my_video_app"
}

Response

Successful Response

namespace_idstring required

Created namespace ID (ns_xxx)

namespace_namestring required

Created namespace name

bucket_idstring required

Created bucket ID (bkt_xxx)

bucket_namestring required

Created bucket name

collection_idstring required

Created collection ID (col_xxx)

collection_namestring required

Created collection name

retriever_idstring required

Created retriever ID (ret_xxx)

retriever_namestring required

Created retriever name

template_idstring required

Scaffold template ID used

statusstring

Status: 'created' for a synchronous (empty) instantiation, 'cloning' when sample data is cloning in the background

task_idstring nullable

Background clone task id when status='cloning' — poll the task, or GET the namespace and read clone_status, to track sample-data progress. Null for synchronous (empty) instantiation.

created_atstring date-time

UTC timestamp of creation

Example response

{
  "bucket_id": "bkt_xyz789",
  "bucket_name": "videos",
  "collection_id": "col_def456",
  "collection_name": "video_embeddings",
  "created_at": "2025-01-15T12:00:00Z",
  "namespace_id": "ns_abc123",
  "namespace_name": "my_video_app",
  "retriever_id": "ret_ghi012",
  "retriever_name": "video_search",
  "status": "created",
  "template_id": "tmpl_scaffold_video_understanding"
}