Skip to main content
Version: v3.0 print this page

Assets

What is a Data Asset?​

Think of a data asset as a valuable piece of data that's been registered in Amorphic. This could be a dataset like structured or unstructured data holding or a glossary. Each data asset comes with metadata, which is like a detailed description covering what the data is about, where it came from, how it's structured, and how it can be used. This makes it easier for user to find the data needed, understand it, and use it effectively for analysis, reporting, and making informed decisions.

Finding Data Assets​

This section helps users find, understand, and use data assets within Amorphic. Users can easily find and explore assets using the Amorphic Catalog feature. View Asset

Key Features​

  • Requesting Access: If user finds an asset that they need but don't have access, user can request permission directly within Amorphic. Request Asset Access If user has access to the undelying resource, they will be able to navigate to it. View Asset Resource
  • Schema: If user has access, they can view the columns (fields) available in datasets. This helps user understand the structure of the data. Asset Schema
  • AI Suggestions: Amorphic uses AI to provide helpful suggestions about the user's data. These suggestions can be found in the Schema section of the asset details.
    • Column Descriptions: Get auto-generated one-liner descriptions that explain what kind of data is stored in each column.
    • Column Classifications: AI classifies the data within a column, suggesting relevant categories from a list of 50+ options.
    • PII (Personally Identifiable Information) Detection: Amorphic detects if a dataset contains PII and classifies it into 250+ categories.
  • Dependent Resources: Identify resources in Amorphic that rely on specific datasets. Asset Dependencies
  • Explore: Access notebooks and studios connected to datasets, allowing user to further investigate the data. Explore

Working with AI Suggestions​

Amorphic offers AI-driven suggestions to help user better understand data. Here's what user needs to know:

  • Availability: AI suggestions are available for datasets with Data Profiling enabled and Target Location: S3Athena, Redshift, Lakeformation, and DynamoDB. Data profiling must be run at least once for the dataset.
  • Reviewing Suggestions: User can approve or decline AI suggestions for column descriptions and classifications at the column or asset level. Approved suggestions are added to the asset schema and become searchable. Declined suggestions are removed.
  • PII Handling: For PII entities, AI continuously monitors data uploads and alerts user upon detection.
  • Important Considerations:
    • To get auto-generated descriptions, ensure the Anthropic Claude V3 Sonnet model is enabled in AWS account. Without it, user can still receive suggestions for PII entities and classifications.
    • AI-generated suggestions may not always be perfect. It's user's responsibility to review and approve or decline them.
    • Once approved or declined, AI suggestions cannot be regenerated. However, user can still manually edit descriptions from the Dataset Details page (under the Profile section).

Auto ML Suggestions

Additional Information​

  • Repair Catalog Metadata: This feature allows users to repair catalog metadata stored in the indexing cluster. It involves deleting the index and re-indexing the data by reading information from the asset metadata tables. It can be accessed from the OS Management tab in the Administration section. Request Asset Access