ML Model Management System Architect#
Model Catalog Data Store
{
"namespace": "/abc_corp/claim",
"model_name": "doc_classification",
"description": "this model is designed for ...",
"owner": "alice@example.com",
"url": "http://abc.corp.com/path/to/the/model/implementation",
# below are details about the model, they can be maintained at:
# 1. stored in the metadata store
# 2. stored in the document file in Git
# 3. stored in the document file and automatically loaded into the metadata store
"data_sources": [
{
"name": "claim history datawarehouse"
"type": "snowflake",
"description": "...",
"owner": "alice@example.com",
"connection": {
"...": "..."
}
},
{
"name": "more data source",
},
]
"inference": {
"input": {
"description": "..."
"schema": {},
},
"output": {
"description": "..."
"schema": {},
},
},
}
Model Registry Metadata Store
{
"namespace": "/abc_corp/claim",
"model_name": "doc_classification",
"model_version": 1,
"description": "this version improves here and there, so that ...",
"artifacts": [
{
"name": "trained",
"uri": "s3://abc_corp/claim/doc_classification/trained/1.tar.gz",
},
{
"name": "static_data",
"uri": "s3://abc_corp/claim/doc_classification/static_data/1.tar.gz",
}
],
"create_at": "2023-01-01T00:00:00.000Z",
"created_by": "arn:aws:iam::123456789012:user/alice",
}
Model Deployment Metadata Store
{
"namespace": "/abc_corp/claim",
"model_name": "doc_classification",
"model_version": 3,
"deployment_id": 1,
"create_at": "2023-01-15T00:00:00.000Z",
"endpoint": "https://abc.execute-api.us-east-1.amazonaws.com/doc_classification_prod",
"environment_variables": [
{"name": "FOO", "value": "BAR"},
{"name": "BAZ", "value": "QUX"},
],
"deployment_method": "self hosted ECS | self hosted EKS | AWS Lambda | AWS SageMaker",
"rolling_strategy": "blue green | canory | rolling",
}
Reference:
Best Machine Learning Model Management Tools That You Need to Know: https://neptune.ai/blog/best-machine-learning-model-management-tools
MLOps: Model management, deployment, and monitoring with Azure Machine Learning: https://learn.microsoft.com/en-us/azure/machine-learning/concept-model-management-and-deployment?view=azureml-api-2