openapi: 3.1.0
info:
  title: Amazon Neptune Neptune ML API
  description: >-
    Neptune ML enables machine learning on graph data using graph neural
    networks. It provides APIs for data processing, model training, model
    transform, and inference endpoint management powered by Amazon SageMaker.
    The ML API endpoints are available on the Neptune DB instance HTTP
    endpoint under the /ml path prefix.
  version: '2024-01-01'
  contact:
    name: Amazon Web Services
    url: https://docs.aws.amazon.com/neptune/latest/userguide/machine-learning.html
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
servers:
- url: https://{cluster-endpoint}:8182
  description: Neptune ML REST endpoint
  variables:
    cluster-endpoint:
      default: your-cluster-endpoint.region.neptune.amazonaws.com
      description: The cluster endpoint DNS name for your Neptune DB cluster
security:
- aws_sigv4: []
tags:
- name: Data Processing
  description: ML data processing job operations
- name: Inference Endpoints
  description: ML inference endpoint management operations
- name: Model Training
  description: ML model training job operations
- name: Model Transform
  description: ML model transform job operations
paths:
  /ml/dataprocessing:
    post:
      operationId: createDataProcessingJob
      summary: Amazon Neptune Create an ML Data Processing Job
      description: >-
        Creates a new Neptune ML data processing job that exports and
        prepares graph data from Neptune for model training using Amazon
        SageMaker. The job reads data from Neptune and writes processed
        output to the specified S3 location.
      tags:
      - Data Processing
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateDataProcessingJobRequest'
      responses:
        '200':
          description: Data processing job created successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobCreatedResponse'
              examples:
                createDataProcessingJob200Example:
                  summary: Default createDataProcessingJob 200 response
                  x-microcks-default: true
                  value:
                    id: neptune-cluster-abc123
        '400':
          description: Bad request - invalid parameters.
        '500':
          description: Internal server error.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
    get:
      operationId: listDataProcessingJobs
      summary: Amazon Neptune List Active Data Processing Jobs
      description: >-
        Returns a list of active Neptune ML data processing job IDs.
      tags:
      - Data Processing
      parameters:
      - name: maxItems
        in: query
        description: Maximum number of items to return (default 10, max 1024).
        schema:
          type: integer
          default: 10
          maximum: 1024
      - name: neptuneIamRoleArn
        in: query
        description: The ARN of the IAM role for Neptune access.
        schema:
          type: string
      responses:
        '200':
          description: Job list retrieved successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobListResponse'
              examples:
                listDataProcessingJobs200Example:
                  summary: Default listDataProcessingJobs 200 response
                  x-microcks-default: true
                  value:
                    ids:
                    - example-value
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
  /ml/dataprocessing/{id}:
    get:
      operationId: getDataProcessingJobStatus
      summary: Amazon Neptune Get Data Processing Job Status
      description: >-
        Returns the status of a Neptune ML data processing job.
      tags:
      - Data Processing
      parameters:
      - name: id
        in: path
        required: true
        description: The unique identifier of the data processing job.
        schema:
          type: string
      - name: neptuneIamRoleArn
        in: query
        description: The ARN of the IAM role for Neptune access.
        schema:
          type: string
      responses:
        '200':
          description: Job status retrieved successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobStatusResponse'
              examples:
                getDataProcessingJobStatus200Example:
                  summary: Default getDataProcessingJobStatus 200 response
                  x-microcks-default: true
                  value:
                    status: available
                    id: neptune-cluster-abc123
                    processingJob:
                      name: my-neptune-cluster
                      arn: arn:aws:neptune:us-east-1:123456789012:db:neptune-cluster-1
                      status: available
                      outputLocation: example-value
                      failureReason: example-value
        '404':
          description: Job not found.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
    delete:
      operationId: stopDataProcessingJob
      summary: Amazon Neptune Stop a Data Processing Job
      description: >-
        Stops a running Neptune ML data processing job. Optionally deletes
        all S3 artifacts when the clean parameter is set to true.
      tags:
      - Data Processing
      parameters:
      - name: id
        in: path
        required: true
        description: The unique identifier of the data processing job.
        schema:
          type: string
      - name: clean
        in: query
        description: Whether to delete all S3 artifacts when stopping.
        schema:
          type: boolean
          default: false
      - name: neptuneIamRoleArn
        in: query
        description: The ARN of the IAM role for Neptune access.
        schema:
          type: string
      responses:
        '200':
          description: Job stopped successfully.
        '404':
          description: Job not found.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
  /ml/modeltraining:
    post:
      operationId: createModelTrainingJob
      summary: Amazon Neptune Create an ML Model Training Job
      description: >-
        Creates a new Neptune ML model training job using Amazon SageMaker.
        The job trains a graph neural network model using the output from
        a completed data processing job. Supports relational graph
        convolutional network (RGCN) and knowledge graph embedding (KGE)
        models, as well as custom training scripts.
      tags:
      - Model Training
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateModelTrainingJobRequest'
      responses:
        '200':
          description: Model training job created successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobCreatedResponse'
              examples:
                createModelTrainingJob200Example:
                  summary: Default createModelTrainingJob 200 response
                  x-microcks-default: true
                  value:
                    id: neptune-cluster-abc123
        '400':
          description: Bad request - invalid parameters.
        '500':
          description: Internal server error.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
    get:
      operationId: listModelTrainingJobs
      summary: Amazon Neptune List Active Model Training Jobs
      description: >-
        Returns a list of active Neptune ML model training job IDs.
      tags:
      - Model Training
      parameters:
      - name: maxItems
        in: query
        description: Maximum number of items to return (default 10, max 1024).
        schema:
          type: integer
          default: 10
          maximum: 1024
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Job list retrieved successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobListResponse'
              examples:
                listModelTrainingJobs200Example:
                  summary: Default listModelTrainingJobs 200 response
                  x-microcks-default: true
                  value:
                    ids:
                    - example-value
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
  /ml/modeltraining/{id}:
    get:
      operationId: getModelTrainingJobStatus
      summary: Amazon Neptune Get Model Training Job Status
      description: >-
        Returns the status of a Neptune ML model training job.
      tags:
      - Model Training
      parameters:
      - name: id
        in: path
        required: true
        description: The unique identifier of the model training job.
        schema:
          type: string
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Job status retrieved successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobStatusResponse'
              examples:
                getModelTrainingJobStatus200Example:
                  summary: Default getModelTrainingJobStatus 200 response
                  x-microcks-default: true
                  value:
                    status: available
                    id: neptune-cluster-abc123
                    processingJob:
                      name: my-neptune-cluster
                      arn: arn:aws:neptune:us-east-1:123456789012:db:neptune-cluster-1
                      status: available
                      outputLocation: example-value
                      failureReason: example-value
        '404':
          description: Job not found.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
    delete:
      operationId: stopModelTrainingJob
      summary: Amazon Neptune Stop a Model Training Job
      description: >-
        Stops a running Neptune ML model training job. Optionally deletes
        all S3 artifacts when the clean parameter is set to true.
      tags:
      - Model Training
      parameters:
      - name: id
        in: path
        required: true
        schema:
          type: string
      - name: clean
        in: query
        schema:
          type: boolean
          default: false
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Job stopped successfully.
        '404':
          description: Job not found.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
  /ml/modeltransform:
    post:
      operationId: createModelTransformJob
      summary: Amazon Neptune Create an ML Model Transform Job
      description: >-
        Creates a new Neptune ML model transform job that generates model
        artifacts for inference. The transform job creates the embeddings
        and other artifacts needed to deploy the model to an inference
        endpoint.
      tags:
      - Model Transform
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateModelTransformJobRequest'
      responses:
        '200':
          description: Model transform job created successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobCreatedResponse'
              examples:
                createModelTransformJob200Example:
                  summary: Default createModelTransformJob 200 response
                  x-microcks-default: true
                  value:
                    id: neptune-cluster-abc123
        '400':
          description: Bad request - invalid parameters.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
    get:
      operationId: listModelTransformJobs
      summary: Amazon Neptune List Active Model Transform Jobs
      description: >-
        Returns a list of active Neptune ML model transform job IDs.
      tags:
      - Model Transform
      parameters:
      - name: maxItems
        in: query
        schema:
          type: integer
          default: 10
          maximum: 1024
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Job list retrieved successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobListResponse'
              examples:
                listModelTransformJobs200Example:
                  summary: Default listModelTransformJobs 200 response
                  x-microcks-default: true
                  value:
                    ids:
                    - example-value
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
  /ml/modeltransform/{id}:
    get:
      operationId: getModelTransformJobStatus
      summary: Amazon Neptune Get Model Transform Job Status
      description: >-
        Returns the status of a Neptune ML model transform job.
      tags:
      - Model Transform
      parameters:
      - name: id
        in: path
        required: true
        schema:
          type: string
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Job status retrieved successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/JobStatusResponse'
              examples:
                getModelTransformJobStatus200Example:
                  summary: Default getModelTransformJobStatus 200 response
                  x-microcks-default: true
                  value:
                    status: available
                    id: neptune-cluster-abc123
                    processingJob:
                      name: my-neptune-cluster
                      arn: arn:aws:neptune:us-east-1:123456789012:db:neptune-cluster-1
                      status: available
                      outputLocation: example-value
                      failureReason: example-value
        '404':
          description: Job not found.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
    delete:
      operationId: stopModelTransformJob
      summary: Amazon Neptune Stop a Model Transform Job
      description: >-
        Stops a running Neptune ML model transform job.
      tags:
      - Model Transform
      parameters:
      - name: id
        in: path
        required: true
        schema:
          type: string
      - name: clean
        in: query
        schema:
          type: boolean
          default: false
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Job stopped successfully.
        '404':
          description: Job not found.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
  /ml/endpoints:
    post:
      operationId: createInferenceEndpoint
      summary: Amazon Neptune Create an ML Inference Endpoint
      description: >-
        Creates a new Neptune ML inference endpoint backed by Amazon
        SageMaker. The endpoint can be used to make real-time predictions
        on graph data using a trained model. Can also be used to update
        an existing endpoint by setting the update parameter to true.
      tags:
      - Inference Endpoints
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateInferenceEndpointRequest'
      responses:
        '200':
          description: Inference endpoint created successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/EndpointCreatedResponse'
              examples:
                createInferenceEndpoint200Example:
                  summary: Default createInferenceEndpoint 200 response
                  x-microcks-default: true
                  value:
                    id: neptune-cluster-abc123
        '400':
          description: Bad request - invalid parameters.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
    get:
      operationId: listInferenceEndpoints
      summary: Amazon Neptune List ML Inference Endpoints
      description: >-
        Returns a list of active Neptune ML inference endpoint IDs.
      tags:
      - Inference Endpoints
      parameters:
      - name: maxItems
        in: query
        schema:
          type: integer
          default: 10
          maximum: 1024
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Endpoint list retrieved successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/EndpointListResponse'
              examples:
                listInferenceEndpoints200Example:
                  summary: Default listInferenceEndpoints 200 response
                  x-microcks-default: true
                  value:
                    ids:
                    - example-value
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
  /ml/endpoints/{id}:
    get:
      operationId: getInferenceEndpointStatus
      summary: Amazon Neptune Get ML Inference Endpoint Status
      description: >-
        Returns the status of a Neptune ML inference endpoint.
      tags:
      - Inference Endpoints
      parameters:
      - name: id
        in: path
        required: true
        description: The unique identifier of the inference endpoint.
        schema:
          type: string
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Endpoint status retrieved successfully.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/EndpointStatusResponse'
              examples:
                getInferenceEndpointStatus200Example:
                  summary: Default getInferenceEndpointStatus 200 response
                  x-microcks-default: true
                  value:
                    status: available
                    id: neptune-cluster-abc123
                    endpoint:
                      name: my-neptune-cluster
                      arn: arn:aws:neptune:us-east-1:123456789012:db:neptune-cluster-1
                      status: available
                    endpointConfig: {}
                    outputLocation: example-value
        '404':
          description: Endpoint not found.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
    delete:
      operationId: deleteInferenceEndpoint
      summary: Amazon Neptune Delete an ML Inference Endpoint
      description: >-
        Deletes a Neptune ML inference endpoint. Optionally cleans up all
        related S3 artifacts.
      tags:
      - Inference Endpoints
      parameters:
      - name: id
        in: path
        required: true
        schema:
          type: string
      - name: clean
        in: query
        description: Whether to delete all related S3 artifacts.
        schema:
          type: boolean
          default: false
      - name: neptuneIamRoleArn
        in: query
        schema:
          type: string
      responses:
        '200':
          description: Endpoint deleted successfully.
        '404':
          description: Endpoint not found.
      x-microcks-operation:
        delay: 0
        dispatcher: FALLBACK
components:
  securitySchemes:
    aws_sigv4:
      type: apiKey
      name: Authorization
      in: header
      description: AWS Signature Version 4 authentication via IAM
  schemas:
    CreateDataProcessingJobRequest:
      type: object
      required:
      - inputDataS3Location
      - processedDataS3Location
      properties:
        id:
          type: string
          description: Unique identifier for the job (auto-generated UUID if omitted).
        inputDataS3Location:
          type: string
          description: S3 URI for the input data.
        processedDataS3Location:
          type: string
          description: S3 URI where processed output is written.
        previousDataProcessingJobId:
          type: string
          description: Job ID of a previous job for incremental processing.
        sagemakerIamRoleArn:
          type: string
          description: IAM role ARN for SageMaker execution.
        neptuneIamRoleArn:
          type: string
          description: IAM role ARN for Neptune access.
        processingInstanceType:
          type: string
          description: ML instance type (default auto-selected ml.r5 type).
        processingInstanceVolumeSizeInGB:
          type: integer
          description: Disk volume size in GB (default 0 means auto-selected).
        processingTimeOutInSeconds:
          type: integer
          description: Timeout in seconds (default 86400, i.e., 1 day).
          default: 86400
        modelType:
          type: string
          description: The type of model to prepare data for.
          enum:
          - heterogeneous
          - kge
        configFileName:
          type: string
          description: The data specification configuration file name.
          default: training-data-configuration.json
        subnets:
          type: array
          description: VPC subnet IDs for SageMaker processing.
          items:
            type: string
        securityGroupIds:
          type: array
          description: VPC security group IDs.
          items:
            type: string
        volumeEncryptionKMSKey:
          type: string
          description: KMS key for storage volume encryption.
        s3OutputEncryptionKMSKey:
          type: string
          description: KMS key for S3 output encryption.
        enableInterContainerTrafficEncryption:
          type: boolean
          description: Whether to enable inter-container traffic encryption.
          default: true
    CreateModelTrainingJobRequest:
      type: object
      required:
      - dataProcessingJobId
      - trainModelS3Location
      properties:
        id:
          type: string
          description: Unique identifier for the job.
        dataProcessingJobId:
          type: string
          description: Job ID of the completed data processing job.
        trainModelS3Location:
          type: string
          description: S3 location for model artifacts output.
        previousModelTrainingJobId:
          type: string
          description: Job ID of a previous training job for incremental training.
        sagemakerIamRoleArn:
          type: string
        neptuneIamRoleArn:
          type: string
        modelName:
          type: string
          description: >-
            The model type to train: rgcn (relational graph convolutional
            network), transe, distmult, rotate, or custom.
          enum:
          - rgcn
          - transe
          - distmult
          - rotate
          - custom
        baseProcessingInstanceType:
          type: string
          description: ML instance type for data preparation step.
        trainingInstanceType:
          type: string
          description: ML instance type for the training step.
          default: ml.p3.2xlarge
        trainingInstanceVolumeSizeInGB:
          type: integer
          description: Disk volume size for training instance in GB.
        trainingTimeOutInSeconds:
          type: integer
          description: Training job timeout in seconds.
          default: 86400
        maxHPONumberOfTrainingJobs:
          type: integer
          description: >-
            Maximum total training jobs for hyperparameter tuning. Minimum
            10 recommended for meaningful results.
          default: 2
        maxHPOParallelTrainingJobs:
          type: integer
          description: Maximum parallel training jobs.
          default: 2
        subnets:
          type: array
          items:
            type: string
        securityGroupIds:
          type: array
          items:
            type: string
        volumeEncryptionKMSKey:
          type: string
        s3OutputEncryptionKMSKey:
          type: string
        enableInterContainerTrafficEncryption:
          type: boolean
          default: true
        enableManagedSpotTraining:
          type: boolean
          description: Whether to use EC2 spot instances for training.
          default: false
        customModelTrainingParameters:
          type: object
          description: Custom model training configuration.
          properties:
            sourceS3DirectoryPath:
              type: string
              description: S3 path to the custom training script directory.
            trainingEntryPointScript:
              type: string
              description: Name of the training entry point script.
            transformEntryPointScript:
              type: string
              description: Name of the transform entry point script.
    CreateModelTransformJobRequest:
      type: object
      required:
      - modelTransformOutputS3Location
      properties:
        id:
          type: string
        dataProcessingJobId:
          type: string
          description: Job ID of the completed data processing job.
        mlModelTrainingJobId:
          type: string
          description: Job ID of the completed model training job.
        trainingJobName:
          type: string
          description: Name of a completed SageMaker training job.
        modelTransformOutputS3Location:
          type: string
          description: S3 location for transform output artifacts.
        sagemakerIamRoleArn:
          type: string
        neptuneIamRoleArn:
          type: string
        baseProcessingInstanceType:
          type: string
        baseProcessingInstanceVolumeSizeInGB:
          type: integer
        subnets:
          type: array
          items:
            type: string
        securityGroupIds:
          type: array
          items:
            type: string
        volumeEncryptionKMSKey:
          type: string
        s3OutputEncryptionKMSKey:
          type: string
        enableInterContainerTrafficEncryption:
          type: boolean
          default: true
        customModelTransformParameters:
          type: object
          properties:
            sourceS3DirectoryPath:
              type: string
            transformEntryPointScript:
              type: string
    CreateInferenceEndpointRequest:
      type: object
      properties:
        id:
          type: string
          description: >-
            Unique identifier for the endpoint (auto-generated timestamped
            name if omitted).
        mlModelTrainingJobId:
          type: string
          description: Job ID from a completed training job.
        mlModelTransformJobId:
          type: string
          description: Job ID from a completed transform job.
        update:
          type: boolean
          description: Whether this is an update to an existing endpoint.
          default: false
        neptuneIamRoleArn:
          type: string
        modelName:
          type: string
          description: The model type.
          enum:
          - rgcn
          - kge
          - transe
          - distmult
          - rotate
        instanceType:
          type: string
          description: ML instance type for the inference endpoint.
          default: ml.m5.xlarge
        instanceCount:
          type: integer
          description: Minimum number of EC2 instances to deploy.
          default: 1
        volumeEncryptionKMSKey:
          type: string
    JobCreatedResponse:
      type: object
      properties:
        id:
          type: string
          description: The unique identifier for the created job.
    JobListResponse:
      type: object
      properties:
        ids:
          type: array
          description: List of job IDs.
          items:
            type: string
    JobStatusResponse:
      type: object
      properties:
        status:
          type: string
          description: The current status of the job.
        id:
          type: string
          description: The job identifier.
        processingJob:
          type: object
          properties:
            name:
              type: string
            arn:
              type: string
            status:
              type: string
            outputLocation:
              type: string
            failureReason:
              type: string
            cloudwatchLogUrl:
              type: string
    EndpointCreatedResponse:
      type: object
      properties:
        id:
          type: string
          description: The unique identifier for the created endpoint.
    EndpointListResponse:
      type: object
      properties:
        ids:
          type: array
          items:
            type: string
    EndpointStatusResponse:
      type: object
      properties:
        status:
          type: string
        id:
          type: string
        endpoint:
          type: object
          properties:
            name:
              type: string
            arn:
              type: string
            status:
              type: string
        endpointConfig:
          type: object
        outputLocation:
          type: string