Every figure remembers where it came from
Why EpistaBase treats a figure as a traceable scientific object, not just an exported image.
A figure is usually where scientific work leaves the system: pasted into slides, sent to a collaborator, or added to a report. That is also where context often gets lost.
EpistaBase treats important figures and tables as records. A published artifact can keep links to:
- the Experiment
- the notebook or analysis session
- successful SDK queries made before publishing
- catalog assets used by the analysis
- source files and parsed tables reachable through lineage
- report blocks that embed the artifact
The user-facing payoff
When someone asks, "Which files and choices made this figure?", the answer should be visible from the figure context, not reconstructed from Slack, downloads, screenshots, and a folder called final_final.
Compliance comes second
Lineage helps with audit and ALCOA+ style expectations, but the docs should open with the working scientist's pain:
- Can I reproduce this result?
- Can my collaborator trust it?
- Can I explain it after the project has moved on?
- Can I tell which raw files and gates were used?
Once those answers are visible, audit language becomes a consequence rather than the headline.