Skip to content

Platform Data Model

Stratum is intentionally split into:

  • a catalog layer for metadata, access, and workflow state
  • an analytical layer for wafer, die, and test-result payloads

This keeps product workflows responsive without treating every analytical query like a transactional concern.

The free stdf.io path is local and browser-first.

The server-side product adds:

  • account and product structure
  • saved lots and snapshots
  • notebooks and collaboration
  • settings and overrides
  • shareable, reviewable analysis artifacts

The useful grains are:

  • lot
  • wafer
  • die
  • test result

Most linked analysis flows are really different projections of those same grains.

If two views disagree, the first question should be whether they are actually looking at the same:

  • lot
  • snapshot
  • wafer scope
  • selection
  • scenario limits

The data model is what makes those contexts propagate coherently.