AWS¶
The sections described below provide techniques for integrating Amazon Web Services with DataRobot.
| Topic | Description |
|---|---|
| Import data from AWS S3 | Importing data from AWS S3 to AI Catalog and creating an ML project. |
| Deploy models on EKS | Deploying and monitor DataRobot models on AWS Elastic Kubernetes Service (EKS) clusters. |
| Path-based routing to PPS on AWS | Using a single IP address for all Portable Prediction Servers through path-based routing. |
| Score Snowflake data on AWS EMR Spark | Scoring Snowflake data via DataRobot models on AWS Elastic Map Reduce (EMR) Spark. |
| Ingest data with AWS Athena | Ingesting AWS Athena and Parquet data for machine learning. |
Lambda¶
| Topic | Description |
|---|---|
| AWS Lambda reporting to MLOps | AWS Lambda serverless reporting of actuals to DataRobot MLOps. |
| Use DataRobot Prime models with AWS Lambda | Using DataRobot Prime* models with AWS Lambda. |
| Use Scoring Code with AWS Lambda | Making predictions using Scoring Code deployed on AWS Lambda. |
* The ability to create new DataRobot Prime models has been removed from the application. This does not affect existing Prime models or deployments.
SageMaker¶
| Topic | Description |
|---|---|
| Deploy models on Sagemaker | Deploying on SageMaker and monitoring with MLOps agents. |
| Use Scoring Code with AWS SageMaker | Making predictions using Scoring Code deployed on AWS SageMaker. |