The article is bit abstract & unclear, but the lists seem to contain some info.

  • What is “active” metadata?
    1. Improved context of metadata powering data discovery
    2. Auto-generated data quality & lineage impact analysis
    3. Auto-classification of sensitive data enabling easy governance & compliance
    4. Making embedded collaboration possible
    5. Orchestration of metadata across platforms
  • requirements:
    • import & export metadata & workflows
    • use ML machine learning
      to recommend job flows, resource allocation, etc.
    • cross-platform metadata analysis
  • features:
    1. always on
    2. don’t just collect metadata, create intelligence
    3. don’t stop at intelligence: drive action
    4. API application programming interface
      -driven embedded collaboration
  • components:
    1. metadata lake: unified (raw & processed) storage repo built on open APIs
    2. Programmable bots: users can create custom DS data science
      / ML machine learning
    3. collaboration plugins: integrations unified by a common metadata layer, seamlessly integrates data tools with each data team’s daily workflow.
    4. Data process automation: easy way to build/deploy/manage workflow automation bots, emulating human decision-making
    5. Reverse metadata: make relevant metadata available to the end-user “rather than in a standalone catalog”