ML Model Service¶

MLOps platforms and model registries - managing machine learning models at scale

Overview¶

The ML Model Service entity represents MLOps platforms and model registries that track, manage, and serve machine learning models. It is the top-level container for ML models and connects to platforms like MLflow, SageMaker, Vertex AI, and other ML lifecycle management tools.

Hierarchy:

graph LR
    A[MLModelService] --> B[MLModel]
    B --> C[Features]

    style A fill:#667eea,color:#fff
    style B fill:#4facfe,color:#fff
    style C fill:#00f2fe,color:#333

Relationships¶

MLModelService has comprehensive relationships with entities across the metadata platform:

graph TD
    subgraph Service Layer
        A[MLModelService<br/>mlflow_prod]
    end

    subgraph ML Models
        A --> B1[MLModel:<br/>churn_predictor]
        A --> B2[MLModel:<br/>fraud_detection]
        A --> B3[MLModel:<br/>recommendation_engine]

        B1 --> C1[Feature:<br/>recency]
        B1 --> C2[Feature:<br/>frequency]
        B1 --> C3[Feature:<br/>monetary]

        B2 --> C4[Feature:<br/>transaction_amount]
        B2 --> C5[Feature:<br/>merchant_category]

        B3 --> C6[Feature:<br/>user_preferences]
        B3 --> C7[Feature:<br/>product_similarity]
    end

    subgraph Ownership
        A -.->|owned by| D1[Team:<br/>Data Science]
        A -.->|owned by| D2[User:<br/>ml.engineer]
    end

    subgraph Governance
        A -.->|in domain| E[Domain:<br/>ML Platform]
        A -.->|has tags| F1[Tag:<br/>Production]
        A -.->|has tags| F2[Tag:<br/>Tier.Gold]
        A -.->|has tags| F3[Tag:<br/>Sensitive]
    end

    subgraph Training Data Lineage
        G1[Table:<br/>customer_features] -.->|trains| B1
        G2[Table:<br/>transaction_history] -.->|trains| B2
        G3[Table:<br/>user_interactions] -.->|trains| B3

        B1 -.->|predictions to| H1[Table:<br/>churn_scores]
        B2 -.->|predictions to| H2[Table:<br/>fraud_scores]
        B3 -.->|predictions to| H3[Table:<br/>recommendations]
    end

    subgraph ML Pipelines
        I1[Pipeline:<br/>feature_engineering] -.->|creates features for| B1
        I2[Pipeline:<br/>model_training] -.->|trains| B1
        I3[Pipeline:<br/>batch_inference] -.->|runs predictions| B1

        J1[Pipeline:<br/>fraud_feature_pipeline] -.->|creates features for| B2
        J2[Pipeline:<br/>fraud_training_pipeline] -.->|trains| B2
    end

    subgraph Model Versions & Metrics
        B1 -.->|has metrics| K1[Accuracy: 92%<br/>AUC: 0.93]
        B2 -.->|has metrics| K2[Precision: 0.85<br/>Recall: 0.88]
        B3 -.->|has metrics| K3[NDCG: 0.78<br/>MAP: 0.72]
    end

    style A fill:#667eea,color:#fff,stroke:#4c51bf,stroke-width:3px
    style B1 fill:#4facfe,color:#fff
    style B2 fill:#4facfe,color:#fff
    style B3 fill:#4facfe,color:#fff
    style C1 fill:#00f2fe,color:#333
    style C2 fill:#00f2fe,color:#333
    style C3 fill:#00f2fe,color:#333
    style C4 fill:#00f2fe,color:#333
    style C5 fill:#00f2fe,color:#333
    style C6 fill:#00f2fe,color:#333
    style C7 fill:#00f2fe,color:#333
    style D1 fill:#43e97b,color:#fff
    style D2 fill:#43e97b,color:#fff
    style E fill:#fa709a,color:#fff
    style F1 fill:#f093fb,color:#fff
    style F2 fill:#f093fb,color:#fff
    style F3 fill:#f093fb,color:#fff
    style G1 fill:#764ba2,color:#fff
    style G2 fill:#764ba2,color:#fff
    style G3 fill:#764ba2,color:#fff
    style H1 fill:#f5576c,color:#fff
    style H2 fill:#f5576c,color:#fff
    style H3 fill:#f5576c,color:#fff
    style I1 fill:#00ac69,color:#fff
    style I2 fill:#00ac69,color:#fff
    style I3 fill:#00ac69,color:#fff
    style J1 fill:#00ac69,color:#fff
    style J2 fill:#00ac69,color:#fff
    style K1 fill:#ffd700,color:#333
    style K2 fill:#ffd700,color:#333
    style K3 fill:#ffd700,color:#333

Relationship Types:

Solid lines (→): Hierarchical containment (Service manages Models, Models have Features)
Dashed lines (-.->): References and associations (ownership, governance, lineage, training, inference)

Child Entities¶

MlModel: ML models managed by this service

Associated Entities¶

Owner: User or team owning this service
Domain: Business domain assignment
Tag: Classification tags
Table: Tables for training data and predictions (via lineage)
Pipeline: Feature engineering, training, and inference pipelines

Schema Specifications¶

View the complete ML Model Service schema in your preferred format:

JSON SchemaRDFJSON-LD

Complete JSON Schema Definition

{
  "$id": "https://open-metadata.org/schema/entity/services/mlmodelService.json",
  "$schema": "http://json-schema.org/draft-07/schema#",
  "title": "MlModelService",
  "description": "MlModel Service Entity, such as MlFlow.",
  "type": "object",
  "javaType": "org.openmetadata.schema.entity.services.MlModelService",
  "javaInterfaces": [
    "org.openmetadata.schema.EntityInterface",
    "org.openmetadata.schema.ServiceEntityInterface"
  ],

  "definitions": {
    "mlModelServiceType": {
      "description": "Type of MlModel service",
      "type": "string",
      "javaInterfaces": ["org.openmetadata.schema.EnumInterface"],
      "enum": ["Mlflow", "Sklearn", "CustomMlModel", "SageMaker", "VertexAI"],
      "javaEnums": [
        {"name": "Mlflow"},
        {"name": "Sklearn"},
        {"name": "CustomMlModel"},
        {"name": "SageMaker"},
        {"name": "VertexAI"}
      ]
    },
    "mlModelConnection": {
      "type": "object",
      "javaType": "org.openmetadata.schema.type.MlModelConnection",
      "description": "MlModel Connection.",
      "javaInterfaces": [
        "org.openmetadata.schema.ServiceConnectionEntityInterface"
      ],
      "properties": {
        "config": {
          "mask": true,
          "oneOf": [
            {"$ref": "./connections/mlmodel/mlflowConnection.json"},
            {"$ref": "./connections/mlmodel/sklearnConnection.json"},
            {"$ref": "./connections/mlmodel/customMlModelConnection.json"},
            {"$ref": "./connections/mlmodel/sageMakerConnection.json"},
            {"$ref": "./connections/mlmodel/vertexaiConnection.json"}
          ]
        }
      },
      "additionalProperties": false
    }
  },

  "properties": {
    "id": {
      "description": "Unique identifier of this pipeline service instance.",
      "$ref": "../../type/basic.json#/definitions/uuid"
    },
    "name": {
      "description": "Name that identifies this pipeline service.",
      "$ref": "../../type/basic.json#/definitions/entityName"
    },
    "fullyQualifiedName": {
      "description": "FullyQualifiedName same as `name`.",
      "$ref": "../../type/basic.json#/definitions/fullyQualifiedEntityName"
    },
    "serviceType": {
      "description": "Type of pipeline service such as Airflow or Prefect...",
      "$ref": "#/definitions/mlModelServiceType"
    },
    "description": {
      "description": "Description of a pipeline service instance.",
      "type": "string"
    },
    "displayName": {
      "description": "Display Name that identifies this pipeline service. It could be title or label from the source services.",
      "type": "string"
    },
    "version": {
      "description": "Metadata version of the entity.",
      "$ref": "../../type/entityHistory.json#/definitions/entityVersion"
    },
    "updatedAt": {
      "description": "Last update time corresponding to the new version of the entity in Unix epoch time milliseconds.",
      "$ref": "../../type/basic.json#/definitions/timestamp"
    },
    "updatedBy": {
      "description": "User who made the update.",
      "type": "string"
    },
    "impersonatedBy": {
      "description": "Bot user that performed the action on behalf of the actual user.",
      "$ref": "../../type/basic.json#/definitions/impersonatedBy"
    },
    "pipelines": {
      "description": "References to pipelines deployed for this pipeline service to extract metadata",
      "$ref": "../../type/entityReferenceList.json"
    },
    "connection": {
      "$ref": "#/definitions/mlModelConnection"
    },
    "testConnectionResult": {
      "description": "Last test connection results for this service",
      "$ref": "connections/testConnectionResult.json"
    },
    "tags": {
      "description": "Tags for this MlModel Service.",
      "type": "array",
      "items": {
        "$ref": "../../type/tagLabel.json"
      },
      "default": []
    },
    "owners": {
      "description": "Owners of this pipeline service.",
      "$ref": "../../type/entityReferenceList.json"
    },
    "href": {
      "description": "Link to the resource corresponding to this pipeline service.",
      "$ref": "../../type/basic.json#/definitions/href"
    },
    "changeDescription": {
      "description": "Change that lead to this version of the entity.",
      "$ref": "../../type/entityHistory.json#/definitions/changeDescription"
    },
    "incrementalChangeDescription": {
      "description": "Change that lead to this version of the entity.",
      "$ref": "../../type/entityHistory.json#/definitions/changeDescription"
    },
    "deleted": {
      "description": "When `true` indicates the entity has been soft deleted.",
      "type": "boolean",
      "default": false
    },
    "dataProducts": {
      "description": "List of data products this entity is part of.",
      "$ref": "../../type/entityReferenceList.json"
    },
    "followers": {
      "description": "Followers of this entity.",
      "$ref": "../../type/entityReferenceList.json"
    },
    "domains": {
      "description": "Domains the MLModel service belongs to.",
      "$ref": "../../type/entityReferenceList.json"
    },
    "ingestionRunner": {
      "description": "The ingestion agent responsible for executing the ingestion pipeline.",
      "$ref": "../../type/entityReference.json"
    }
  },

  "required": ["id", "name", "serviceType"],
  "additionalProperties": false
}

View Full JSON Schema →

RDF/OWL Ontology Definition

@prefix om: <https://open-metadata.org/schema/> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix owl: <http://www.w3.org/2002/07/owl#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .

# MLModelService Class Definition
om:MlModelService a owl:Class ;
    rdfs:subClassOf om:Service ;
    rdfs:label "ML Model Service" ;
    rdfs:comment "MlModel Service Entity, such as MlFlow." ;
    om:hierarchyLevel 1 .

# Properties
om:serviceName a owl:DatatypeProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range xsd:string ;
    rdfs:label "name" ;
    rdfs:comment "Name that identifies this pipeline service." .

om:serviceType a owl:DatatypeProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:MlModelServiceType ;
    rdfs:label "serviceType" ;
    rdfs:comment "Type of MlModel service" .

om:description a owl:DatatypeProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range xsd:string ;
    rdfs:label "description" ;
    rdfs:comment "Description of a pipeline service instance." .

om:displayName a owl:DatatypeProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range xsd:string ;
    rdfs:label "displayName" ;
    rdfs:comment "Display Name that identifies this pipeline service." .

om:hasMlModel a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:MlModel ;
    rdfs:label "hasMlModel" ;
    rdfs:comment "Models managed by this service" .

om:owners a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:EntityReferenceList ;
    rdfs:label "owners" ;
    rdfs:comment "Owners of this pipeline service." .

om:domains a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:EntityReferenceList ;
    rdfs:label "domains" ;
    rdfs:comment "Domains the MLModel service belongs to." .

om:tags a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:TagLabel ;
    rdfs:label "tags" ;
    rdfs:comment "Tags for this MlModel Service." .

om:followers a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:EntityReferenceList ;
    rdfs:label "followers" ;
    rdfs:comment "Followers of this entity." .

om:dataProducts a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:EntityReferenceList ;
    rdfs:label "dataProducts" ;
    rdfs:comment "List of data products this entity is part of." .

om:connection a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:MlModelConnection ;
    rdfs:label "connection" ;
    rdfs:comment "MlModel Connection." .

om:pipelines a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:EntityReferenceList ;
    rdfs:label "pipelines" ;
    rdfs:comment "References to pipelines deployed for this pipeline service to extract metadata" .

om:ingestionRunner a owl:ObjectProperty ;
    rdfs:domain om:MlModelService ;
    rdfs:range om:EntityReference ;
    rdfs:label "ingestionRunner" ;
    rdfs:comment "The ingestion agent responsible for executing the ingestion pipeline." .

# Service Type Enumeration
om:MlModelServiceType a owl:Class ;
    owl:oneOf (
        om:Mlflow
        om:Sklearn
        om:CustomMlModel
        om:SageMaker
        om:VertexAI
    ) .

# Example Instance
ex:mlflowProduction a om:MlModelService ;
    om:serviceName "mlflow_prod" ;
    om:fullyQualifiedName "mlflow_prod" ;
    om:displayName "MLflow Production" ;
    om:serviceType om:Mlflow ;
    om:owners ex:dataScienceTeam ;
    om:tags ex:tierProduction ;
    om:domains ex:mlPlatformDomain ;
    om:hasMlModel ex:churnPredictor ;
    om:hasMlModel ex:fraudDetection .

View Full RDF Ontology →

JSON-LD Context and Example

{
  "@context": {
    "@vocab": "https://open-metadata.org/schema/",
    "om": "https://open-metadata.org/schema/",
    "rdfs": "http://www.w3.org/2000/01/rdf-schema#",
    "xsd": "http://www.w3.org/2001/XMLSchema#",

    "MlModelService": "om:MlModelService",
    "name": {
      "@id": "om:serviceName",
      "@type": "xsd:string"
    },
    "fullyQualifiedName": {
      "@id": "om:fullyQualifiedName",
      "@type": "xsd:string"
    },
    "displayName": {
      "@id": "om:displayName",
      "@type": "xsd:string"
    },
    "description": {
      "@id": "om:description",
      "@type": "xsd:string"
    },
    "serviceType": {
      "@id": "om:serviceType",
      "@type": "@vocab"
    },
    "connection": {
      "@id": "om:connection",
      "@type": "@id"
    },
    "owners": {
      "@id": "om:owners",
      "@type": "@id",
      "@container": "@set"
    },
    "domains": {
      "@id": "om:domains",
      "@type": "@id",
      "@container": "@set"
    },
    "tags": {
      "@id": "om:tags",
      "@type": "@id",
      "@container": "@set"
    },
    "followers": {
      "@id": "om:followers",
      "@type": "@id",
      "@container": "@set"
    },
    "dataProducts": {
      "@id": "om:dataProducts",
      "@type": "@id",
      "@container": "@set"
    },
    "pipelines": {
      "@id": "om:pipelines",
      "@type": "@id",
      "@container": "@set"
    },
    "ingestionRunner": {
      "@id": "om:ingestionRunner",
      "@type": "@id"
    }
  }
}

Example JSON-LD Instance:

{
  "@context": "https://open-metadata.org/context/mlmodelService.jsonld",
  "@type": "MlModelService",
  "@id": "https://example.com/services/mlflow_prod",

  "name": "mlflow_prod",
  "fullyQualifiedName": "mlflow_prod",
  "displayName": "MLflow Production",
  "description": "Production MLflow instance for model tracking and registry",
  "serviceType": "Mlflow",

  "connection": {
    "config": {
      "type": "Mlflow",
      "trackingUri": "https://mlflow.company.com",
      "registryUri": "https://mlflow.company.com"
    }
  },

  "owners": [
    {
      "@id": "https://example.com/teams/data-science",
      "@type": "Team",
      "name": "data-science",
      "displayName": "Data Science Team"
    }
  ],

  "domains": [
    {
      "@id": "https://example.com/domains/AI-ML",
      "@type": "Domain",
      "name": "AI-ML"
    }
  ],

  "tags": [
    {
      "@id": "https://open-metadata.org/tags/Tier/Production",
      "tagFQN": "Tier.Production"
    },
    {
      "@id": "https://open-metadata.org/tags/Environment/Prod",
      "tagFQN": "Environment.Prod"
    }
  ],

  "followers": [],
  "dataProducts": []
}

View Full JSON-LD Context →

Use Cases¶

Connect to MLflow, SageMaker, Vertex AI, and other ML platforms
Discover and catalog ML models across different registries
Track model ownership and governance by ML platform
Apply environment tags (production, staging, development)
Organize models by team and domain
Monitor model deployment across platforms
Ensure compliance for AI/ML systems
Enable cross-platform model lineage

JSON Schema Specification¶

Core Properties¶

`id` (uuid)¶

Type: string (UUID format) Required: Yes (system-generated) Description: Unique identifier for this ML model service instance

{
  "id": "7a8b9c0d-1e2f-3a4b-5c6d-7e8f9a0b1c2d"
}

`name` (entityName)¶

Type: string Required: Yes Pattern: ^[^.]*$ (no dots allowed) Min Length: 1 Max Length: 256 Description: Name of the ML model service

{
  "name": "mlflow_prod"
}

`fullyQualifiedName` (fullyQualifiedEntityName)¶

Type: string Required: Yes (system-generated) Description: For services, this is the same as name

{
  "fullyQualifiedName": "mlflow_prod"
}

`displayName`¶

Type: string Required: No Description: Human-readable display name

{
  "displayName": "MLflow Production Environment"
}

`description` (markdown)¶

Type: string (Markdown format) Required: No Description: Rich text description of the service's purpose and usage

{
  "description": "# MLflow Production\n\nProduction MLflow instance for tracking experiments, registering models, and serving predictions.\n\n## Usage\n- All production models must be registered here\n- Experiment tracking for data science team\n- Model deployment via SageMaker integration"
}

Service Configuration¶

`serviceType` (mlModelServiceType enum)¶

Type: string enum Required: Yes Allowed Values:

Mlflow - Open-source ML lifecycle management
Sklearn - Scikit-learn models
CustomMlModel - Custom ML platform
SageMaker - AWS SageMaker model registry
VertexAI - Google Cloud Vertex AI

{
  "serviceType": "Mlflow"
}

`connection` (MLModelConnection)¶

Type: object Required: Yes Description: Connection configuration specific to the service type

MLflow Connection:

{
  "connection": {
    "config": {
      "type": "MLflow",
      "trackingUri": "https://mlflow.company.com",
      "registryUri": "https://mlflow.company.com",
      "supportsMetadataExtraction": true
    }
  }
}

AWS SageMaker Connection:

{
  "connection": {
    "config": {
      "type": "SageMaker",
      "awsRegion": "us-west-2",
      "awsAccessKeyId": "${AWS_ACCESS_KEY_ID}",
      "awsSecretAccessKey": "${AWS_SECRET_ACCESS_KEY}",
      "awsSessionToken": "${AWS_SESSION_TOKEN}",
      "endpointURL": "https://api.sagemaker.us-west-2.amazonaws.com"
    }
  }
}

Google Vertex AI Connection:

{
  "connection": {
    "config": {
      "type": "VertexAI",
      "gcpProjectId": "my-ml-project",
      "gcpRegion": "us-central1",
      "credentials": {
        "gcpConfig": {
          "type": "service_account",
          "projectId": "my-ml-project",
          "privateKeyId": "key-id",
          "privateKey": "-----BEGIN PRIVATE KEY-----\n...\n-----END PRIVATE KEY-----\n",
          "clientEmail": "ml-service@my-ml-project.iam.gserviceaccount.com"
        }
      }
    }
  }
}

Azure ML Connection:

{
  "connection": {
    "config": {
      "type": "AzureML",
      "subscriptionId": "azure-subscription-id",
      "resourceGroup": "ml-resource-group",
      "workspaceName": "ml-workspace",
      "tenantId": "azure-tenant-id",
      "clientId": "azure-client-id",
      "clientSecret": "${AZURE_CLIENT_SECRET}"
    }
  }
}

Governance Properties¶

`owners` (EntityReferenceList)¶

Type: array (EntityReferenceList) Required: No Description: Owners of this pipeline service

{
  "owners": [
    {
      "id": "9d8e7f6a-5b4c-3d2e-1f0a-9b8c7d6e5f4a",
      "type": "team",
      "name": "data-science",
      "displayName": "Data Science Team"
    }
  ]
}

`domains` (EntityReferenceList)¶

Type: array (EntityReferenceList) Required: No Description: Domains the MLModel service belongs to

{
  "domains": [
    {
      "id": "a1b2c3d4-e5f6-7a8b-9c0d-1e2f3a4b5c6d",
      "type": "domain",
      "name": "AI-ML",
      "fullyQualifiedName": "AI-ML"
    }
  ]
}

`tags[]` (TagLabel[])¶

Type: array Required: No Default: [] Description: Tags for this MlModel Service

{
  "tags": [
    {
      "tagFQN": "Tier.Production",
      "description": "Production ML infrastructure",
      "source": "Classification",
      "labelType": "Manual",
      "state": "Confirmed"
    },
    {
      "tagFQN": "Environment.Prod",
      "source": "Classification",
      "labelType": "Manual",
      "state": "Confirmed"
    }
  ]
}

`followers` (EntityReferenceList)¶

Type: array (EntityReferenceList) Required: No Description: Followers of this entity

{
  "followers": [
    {
      "id": "1a2b3c4d-5e6f-7a8b-9c0d-1e2f3a4b5c6d",
      "type": "user",
      "name": "john.doe",
      "displayName": "John Doe"
    }
  ]
}

`dataProducts` (EntityReferenceList)¶

Type: array (EntityReferenceList) Required: No Description: List of data products this entity is part of

{
  "dataProducts": [
    {
      "id": "2b3c4d5e-6f7a-8b9c-0d1e-2f3a4b5c6d7e",
      "type": "dataProduct",
      "name": "ml-insights",
      "fullyQualifiedName": "ml-insights"
    }
  ]
}

`pipelines` (EntityReferenceList)¶

Type: array (EntityReferenceList) Required: No Description: References to pipelines deployed for this pipeline service to extract metadata

{
  "pipelines": [
    {
      "id": "3c4d5e6f-7a8b-9c0d-1e2f-3a4b5c6d7e8f",
      "type": "pipeline",
      "name": "mlflow_ingestion",
      "fullyQualifiedName": "mlflow_prod.mlflow_ingestion"
    }
  ]
}

`ingestionRunner` (EntityReference)¶

Type: object (EntityReference) Required: No Description: The ingestion agent responsible for executing the ingestion pipeline

{
  "ingestionRunner": {
    "id": "4d5e6f7a-8b9c-0d1e-2f3a-4b5c6d7e8f9a",
    "type": "ingestionPipeline",
    "name": "openmetadata-agent"
  }
}

`href` (href)¶

Type: string (URI) Required: No Description: Link to the resource corresponding to this pipeline service

{
  "href": "http://localhost:8585/api/v1/services/mlmodelServices/7a8b9c0d-1e2f-3a4b-5c6d-7e8f9a0b1c2d"
}

`testConnectionResult` (TestConnectionResult)¶

Type: object Required: No Description: Last test connection results for this service

{
  "testConnectionResult": {
    "status": "Successful",
    "lastUpdatedAt": 1704240000000
  }
}

`deleted` (boolean)¶

Type: boolean Required: No Default: false Description: When true indicates the entity has been soft deleted

{
  "deleted": false
}

`changeDescription` (ChangeDescription)¶

Type: object Required: No Description: Change that lead to this version of the entity

{
  "changeDescription": {
    "fieldsAdded": [],
    "fieldsUpdated": [
      {
        "name": "tags",
        "oldValue": [],
        "newValue": [{"tagFQN": "Tier.Production"}]
      }
    ],
    "fieldsDeleted": [],
    "previousVersion": 1.0
  }
}

`incrementalChangeDescription` (ChangeDescription)¶

Type: object Required: No Description: Change that lead to this version of the entity

{
  "incrementalChangeDescription": {
    "fieldsAdded": [],
    "fieldsUpdated": [],
    "fieldsDeleted": [],
    "previousVersion": 1.1
  }
}

`impersonatedBy` (impersonatedBy)¶

Type: string Required: No Description: Bot user that performed the action on behalf of the actual user

{
  "impersonatedBy": "ingestion-bot"
}

Versioning Properties¶

`version` (entityVersion)¶

Type: number Required: Yes (system-managed) Description: Metadata version number, incremented on changes

{
  "version": 1.2
}

`updatedAt` (timestamp)¶

Type: integer (Unix epoch milliseconds) Required: Yes (system-managed) Description: Last update timestamp

{
  "updatedAt": 1704240000000
}

`updatedBy` (string)¶

Type: string Required: Yes (system-managed) Description: User who made the update

{
  "updatedBy": "admin"
}

Complete Example¶

MLflow Production Service¶

{
  "id": "7a8b9c0d-1e2f-3a4b-5c6d-7e8f9a0b1c2d",
  "name": "mlflow_prod",
  "fullyQualifiedName": "mlflow_prod",
  "displayName": "MLflow Production",
  "description": "Production MLflow tracking server and model registry for tracking experiments, registering models, and serving predictions.",
  "serviceType": "Mlflow",
  "connection": {
    "config": {
      "type": "Mlflow",
      "trackingUri": "https://mlflow.company.com",
      "registryUri": "https://mlflow.company.com"
    }
  },
  "owners": [
    {
      "id": "9d8e7f6a-5b4c-3d2e-1f0a-9b8c7d6e5f4a",
      "type": "team",
      "name": "data-science",
      "displayName": "Data Science Team"
    }
  ],
  "domains": [
    {
      "id": "a1b2c3d4-e5f6-7a8b-9c0d-1e2f3a4b5c6d",
      "type": "domain",
      "name": "AI-ML"
    }
  ],
  "tags": [
    {"tagFQN": "Tier.Production"},
    {"tagFQN": "Environment.Prod"}
  ],
  "followers": [],
  "dataProducts": [],
  "deleted": false,
  "version": 1.0,
  "updatedAt": 1704240000000,
  "updatedBy": "admin"
}

SageMaker Service¶

{
  "id": "2b3c4d5e-6f7a-8b9c-0d1e-2f3a4b5c6d7e",
  "name": "sagemaker_models",
  "fullyQualifiedName": "sagemaker_models",
  "displayName": "AWS SageMaker Production",
  "description": "Production SageMaker instance for model deployment and serving",
  "serviceType": "SageMaker",
  "connection": {
    "config": {
      "type": "SageMaker",
      "awsRegion": "us-west-2",
      "awsAccessKeyId": "${AWS_ACCESS_KEY_ID}",
      "awsSecretAccessKey": "${AWS_SECRET_ACCESS_KEY}"
    }
  },
  "owners": [
    {
      "id": "9d8e7f6a-5b4c-3d2e-1f0a-9b8c7d6e5f4a",
      "type": "team",
      "name": "ml-ops",
      "displayName": "ML Operations Team"
    }
  ],
  "tags": [
    {"tagFQN": "Tier.Production"},
    {"tagFQN": "Cloud.AWS"}
  ],
  "deleted": false,
  "version": 1.0,
  "updatedAt": 1704240000000,
  "updatedBy": "admin"
}

RDF Representation¶

Ontology Class¶

@prefix om: <https://open-metadata.org/schema/> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix owl: <http://www.w3.org/2002/07/owl#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .

om:MlModelService a owl:Class ;
    rdfs:subClassOf om:Service ;
    rdfs:label "ML Model Service" ;
    rdfs:comment "MlModel Service Entity, such as MlFlow." ;
    om:hasProperties [
        om:name "string" ;
        om:serviceType "MlModelServiceType" ;
        om:connection "MlModelConnection" ;
        om:owners "EntityReferenceList" ;
        om:domains "EntityReferenceList" ;
        om:tags "TagLabel[]" ;
    ] .

Instance Example¶

@prefix om: <https://open-metadata.org/schema/> .
@prefix ex: <https://example.com/services/> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .

ex:mlflow_prod a om:MlModelService ;
    om:serviceName "mlflow_prod" ;
    om:fullyQualifiedName "mlflow_prod" ;
    om:displayName "MLflow Production" ;
    om:serviceType om:Mlflow ;
    om:owners ex:data_science_team ;
    om:domains ex:ai_ml_domain ;
    om:tags ex:tier_production ;
    om:tags ex:environment_prod ;
    om:hasMlModel ex:churn_predictor ;
    om:hasMlModel ex:fraud_detection ;
    om:deleted "false"^^xsd:boolean .

JSON-LD Context¶

{
  "@context": {
    "@vocab": "https://open-metadata.org/schema/",
    "om": "https://open-metadata.org/schema/",
    "MlModelService": "om:MlModelService",
    "name": "om:serviceName",
    "fullyQualifiedName": "om:fullyQualifiedName",
    "displayName": "om:displayName",
    "description": "om:description",
    "serviceType": {
      "@id": "om:serviceType",
      "@type": "@vocab"
    },
    "connection": {
      "@id": "om:connection",
      "@type": "@id"
    },
    "owners": {
      "@id": "om:owners",
      "@type": "@id",
      "@container": "@set"
    },
    "domains": {
      "@id": "om:domains",
      "@type": "@id",
      "@container": "@set"
    },
    "tags": {
      "@id": "om:tags",
      "@type": "@id",
      "@container": "@set"
    },
    "followers": {
      "@id": "om:followers",
      "@type": "@id",
      "@container": "@set"
    },
    "deleted": {
      "@id": "om:deleted",
      "@type": "xsd:boolean"
    }
  }
}

JSON-LD Example¶

{
  "@context": "https://open-metadata.org/context/mlmodelService.jsonld",
  "@type": "MlModelService",
  "@id": "https://example.com/services/mlflow_prod",
  "name": "mlflow_prod",
  "displayName": "MLflow Production",
  "description": "Production MLflow tracking server and model registry",
  "serviceType": "Mlflow",
  "owners": [
    {
      "@id": "https://example.com/teams/data-science",
      "@type": "Team",
      "name": "data-science"
    }
  ],
  "domains": [
    {
      "@id": "https://example.com/domains/AI-ML",
      "@type": "Domain",
      "name": "AI-ML"
    }
  ],
  "tags": [
    {"@id": "https://open-metadata.org/tags/Tier/Production"}
  ],
  "deleted": false
}

Custom Properties¶

This entity supports custom properties through the extension field. Common custom properties include:

Data Classification: Sensitivity level
Cost Center: Billing allocation
Retention Period: Data retention requirements
Application Owner: Owning application/team

See Custom Properties for details on defining and using custom properties.

API Operations¶

Create ML Model Service¶

POST /api/v1/services/mlmodelServices
Content-Type: application/json

{
  "name": "mlflow_prod",
  "serviceType": "MLflow",
  "connection": {
    "config": {
      "type": "MLflow",
      "trackingUri": "https://mlflow.company.com"
    }
  }
}

Get ML Model Service¶

GET /api/v1/services/mlmodelServices/name/mlflow_prod?fields=owner,tags,domain

Update ML Model Service¶

PATCH /api/v1/services/mlmodelServices/{id}
Content-Type: application/json-patch+json

[
  {
    "op": "add",
    "path": "/tags/-",
    "value": {"tagFQN": "Tier.Production"}
  }
]

Test Connection¶

POST /api/v1/services/mlmodelServices/testConnection
Content-Type: application/json

{
  "serviceType": "MLflow",
  "connection": {
    "config": {
      "type": "MLflow",
      "trackingUri": "https://mlflow.company.com"
    }
  }
}

List Models in Service¶

GET /api/v1/mlmodels?service=mlflow_prod&fields=algorithm,owner

ML Model - ML model specification
ML Overview - ML assets overview
Services - Service entity patterns
Lineage - Model lineage tracking

ML Model Service¶

Overview¶

Relationships¶

Child Entities¶

Associated Entities¶

Schema Specifications¶

Use Cases¶

JSON Schema Specification¶

Core Properties¶

id (uuid)¶

name (entityName)¶

fullyQualifiedName (fullyQualifiedEntityName)¶

displayName¶

description (markdown)¶

Service Configuration¶

serviceType (mlModelServiceType enum)¶

connection (MLModelConnection)¶