Select Language:
If you’re looking to make your AWS Glue Data Catalog more informative and easier to understand, you can use Amazon SageMaker Unified Studio’s AI tools to automatically generate metadata for your tables. This process can help you save time and make your data more discoverable without manual effort.
To start, head over to your SageMaker Unified Studio and select the Data tab for your project. From the side menu, click on Data sources and choose the data source connected to your Glue Data Catalog database. Then, go to the details section and turn on the setting called Automated Business Name Generation. Once enabled, SageMaker will use Amazon Bedrock’s advanced language models to automatically create descriptive business names for your tables.
If you prefer to generate these names when creating new assets, you can do so by adjusting the API settings. By setting the businessNameGeneration flag within predictionConfiguration during asset creation, SageMaker will automatically generate suitable names for you.
Apart from business names, SageMaker Unified Studio can also help create other useful metadata like business descriptions, summaries, and glossary terms. These AI-suggested details help clarify the purpose of each table and make your catalogs more comprehensive. You can review these suggestions and choose to accept, edit, or reject them based on your needs.
For more detailed metadata, you might want to create custom forms within SageMaker to add structured business context to your assets. This extra step ensures your data catalog is rich in meaningful information to support data discovery and analysis.
Learn more about this process and how to customize metadata forms through official AWS guides linked below:
– Using machine learning and generative AI in Amazon SageMaker Unified Studio
– Creating a metadata form in Amazon SageMaker Unified Studio
ChatGPT
Perplexity
Gemini AI
Grok AI
