Vectorize Pipelines API

openapi: 3.0.1
info:
  title: Vectorize AI Platform Connectors Pipelines API
  description: 'The Vectorize API lets developers programmatically build and operate RAG (retrieval-augmented generation) pipelines: create / start / stop / delete pipelines, run retrieval against a pipeline''s vector index, manage source, destination, and AI-platform connectors, upload files, extract documents with Iris, and run deep research. All paths are organization-scoped under /org/{organizationId}. Authentication uses a Bearer access token.'
  termsOfService: https://vectorize.io/terms
  contact:
    name: Vectorize Support
    url: https://vectorize.io
  version: 0.1.2
servers:
- url: https://api.vectorize.io/v1
  description: Vectorize API
security:
- bearerAuth: []
tags:
- name: Pipelines
  description: Create and operate RAG pipelines, retrieval, and deep research.
paths:
  /org/{organizationId}/pipelines:
    post:
      operationId: createPipeline
      tags:
      - Pipelines
      summary: Create a new pipeline
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/PipelineConfigurationSchema'
      responses:
        '200':
          description: Pipeline created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/CreatePipelineResponse'
        '400':
          $ref: '#/components/responses/Error'
        '401':
          $ref: '#/components/responses/Error'
        '500':
          $ref: '#/components/responses/Error'
    get:
      operationId: getPipelines
      tags:
      - Pipelines
      summary: List pipelines for an organization
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      responses:
        '200':
          description: A list of pipelines.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/GetPipelinesResponse'
        '401':
          $ref: '#/components/responses/Error'
        '500':
          $ref: '#/components/responses/Error'
  /org/{organizationId}/pipelines/{pipelineId}:
    get:
      operationId: getPipeline
      tags:
      - Pipelines
      summary: Retrieve a pipeline
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      responses:
        '200':
          description: The pipeline.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/PipelineListSummary'
        '404':
          $ref: '#/components/responses/Error'
    delete:
      operationId: deletePipeline
      tags:
      - Pipelines
      summary: Delete a pipeline
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      responses:
        '200':
          description: Pipeline deleted.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/MessageResponse'
        '404':
          $ref: '#/components/responses/Error'
  /org/{organizationId}/pipelines/{pipelineId}/start:
    post:
      operationId: startPipeline
      tags:
      - Pipelines
      summary: Start a pipeline
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      responses:
        '200':
          description: Pipeline started.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/MessageResponse'
        '404':
          $ref: '#/components/responses/Error'
  /org/{organizationId}/pipelines/{pipelineId}/stop:
    post:
      operationId: stopPipeline
      tags:
      - Pipelines
      summary: Stop a pipeline
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      responses:
        '200':
          description: Pipeline stopped.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/MessageResponse'
        '404':
          $ref: '#/components/responses/Error'
  /org/{organizationId}/pipelines/{pipelineId}/retrieval:
    post:
      operationId: retrieveDocuments
      tags:
      - Pipelines
      summary: Retrieve documents from a pipeline
      description: Vectorizes the input question, runs a k-ANN search against the pipeline's vector index, optionally reranks, and returns the most relevant document chunks ordered by relevancy / similarity score.
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/RetrieveDocumentsRequest'
      responses:
        '200':
          description: Retrieved documents.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/RetrieveDocumentsResponse'
        '400':
          $ref: '#/components/responses/Error'
        '404':
          $ref: '#/components/responses/Error'
        '500':
          $ref: '#/components/responses/Error'
  /org/{organizationId}/pipelines/{pipelineId}/deep-research:
    post:
      operationId: startDeepResearch
      tags:
      - Pipelines
      summary: Start a deep research
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/StartDeepResearchRequest'
      responses:
        '200':
          description: Deep research started.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/StartDeepResearchResponse'
        '400':
          $ref: '#/components/responses/Error'
        '404':
          $ref: '#/components/responses/Error'
  /org/{organizationId}/pipelines/{pipelineId}/deep-research/{researchId}:
    get:
      operationId: getDeepResearchResult
      tags:
      - Pipelines
      summary: Get a deep research result
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      - name: researchId
        in: path
        required: true
        schema:
          type: string
      responses:
        '200':
          description: The deep research result.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/DeepResearchResultResponse'
        '404':
          $ref: '#/components/responses/Error'
  /org/{organizationId}/pipelines/{pipelineId}/events:
    get:
      operationId: getPipelineEvents
      tags:
      - Pipelines
      summary: Get pipeline events
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      responses:
        '200':
          description: Pipeline events.
          content:
            application/json:
              schema:
                type: object
                additionalProperties: true
        '404':
          $ref: '#/components/responses/Error'
  /org/{organizationId}/pipelines/{pipelineId}/metrics:
    get:
      operationId: getPipelineMetrics
      tags:
      - Pipelines
      summary: Get pipeline metrics
      parameters:
      - $ref: '#/components/parameters/OrganizationId'
      - $ref: '#/components/parameters/PipelineId'
      responses:
        '200':
          description: Pipeline metrics.
          content:
            application/json:
              schema:
                type: object
                additionalProperties: true
        '404':
          $ref: '#/components/responses/Error'
components:
  schemas:
    ErrorResponse:
      type: object
      properties:
        error:
          type: string
        details:
          type: string
        failedUpdates:
          type: array
          items:
            type: string
        successfulUpdates:
          type: array
          items:
            type: string
      required:
      - error
    CreatePipelineResponse:
      type: object
      properties:
        message:
          type: string
        data:
          type: object
          properties:
            id:
              type: string
          required:
          - id
      required:
      - message
      - data
    GetPipelinesResponse:
      type: object
      properties:
        message:
          type: string
        data:
          type: array
          items:
            $ref: '#/components/schemas/PipelineListSummary'
      required:
      - message
      - data
    Document:
      type: object
      properties:
        relevancy:
          type: number
        id:
          type: string
        text:
          type: string
        chunk_id:
          type: string
        total_chunks:
          type: string
        origin:
          type: string
        origin_id:
          type: string
        similarity:
          type: number
        source:
          type: string
        unique_source:
          type: string
        source_display_name:
          type: string
        pipeline_id:
          type: string
        org_id:
          type: string
      required:
      - relevancy
      - id
      - text
      - chunk_id
      - total_chunks
      - origin
      - origin_id
      - similarity
      - source
      - unique_source
      - source_display_name
      additionalProperties: true
    MessageResponse:
      type: object
      properties:
        message:
          type: string
      required:
      - message
    StartDeepResearchResponse:
      type: object
      properties:
        researchId:
          type: string
      required:
      - researchId
    PipelineListSummary:
      type: object
      properties:
        id:
          type: string
        name:
          type: string
        documentCount:
          type: number
        sourceConnectorAuthIds:
          type: array
          items:
            type: string
        destinationConnectorAuthIds:
          type: array
          items:
            type: string
        aiPlatformAuthIds:
          type: array
          items:
            type: string
        sourceConnectorTypes:
          type: array
          items:
            type: string
        destinationConnectorTypes:
          type: array
          items:
            type: string
        aiPlatformTypes:
          type: array
          items:
            type: string
        createdAt:
          type: string
          nullable: true
      additionalProperties: true
    StartDeepResearchRequest:
      type: object
      properties:
        query:
          type: string
        webSearch:
          type: boolean
          default: false
        schema:
          type: string
        n8n:
          type: object
          description: Optional n8n workflow configuration.
          additionalProperties: true
      required:
      - query
    RetrieveDocumentsResponse:
      type: object
      properties:
        question:
          type: string
        documents:
          type: array
          items:
            $ref: '#/components/schemas/Document'
        average_relevancy:
          type: number
        ndcg:
          type: number
      required:
      - question
      - documents
      - average_relevancy
      - ndcg
    DeepResearchResultResponse:
      type: object
      properties:
        ready:
          type: boolean
        data:
          type: object
          description: The generated deep research report when ready.
          additionalProperties: true
      required:
      - ready
    PipelineConfigurationSchema:
      type: object
      properties:
        sourceConnectors:
          type: array
          minItems: 1
          items:
            type: object
            description: A pipeline source connector reference and its configuration.
            additionalProperties: true
        destinationConnector:
          type: object
          description: The destination (vector database) connector for the pipeline.
          additionalProperties: true
        aiPlatformConnector:
          type: object
          description: The AI platform (embedding / model) connector for the pipeline.
          additionalProperties: true
        pipelineName:
          type: string
          minLength: 1
        schedule:
          type: object
          description: Refresh schedule for the pipeline.
          additionalProperties: true
      required:
      - sourceConnectors
      - destinationConnector
      - aiPlatformConnector
      - pipelineName
      - schedule
    RetrieveDocumentsRequest:
      type: object
      properties:
        question:
          type: string
        numResults:
          type: number
          minimum: 1
        rerank:
          type: boolean
          default: true
        metadata-filters:
          type: array
          items:
            type: object
            additionalProperties:
              nullable: true
        context:
          type: object
          description: Optional context object used for query rewriting.
          additionalProperties: true
        advanced-query:
          type: object
          description: Optional advanced-query configuration.
          additionalProperties: true
      required:
      - question
      - numResults
  parameters:
    PipelineId:
      name: pipelineId
      in: path
      required: true
      schema:
        type: string
    OrganizationId:
      name: organizationId
      in: path
      required: true
      description: Vectorize organization identifier.
      schema:
        type: string
  responses:
    Error:
      description: Error response.
      content:
        application/json:
          schema:
            $ref: '#/components/schemas/ErrorResponse'
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      bearerFormat: JWT
      description: Vectorize API access token passed as a Bearer token.