Assays that explain themselves

A result you cannot reproduce is a rumour. Most of the friction in reproducing a wet-lab assay is not the biology — it is the missing context: which reagent lot, which calibration, which deviation from protocol went unrecorded.

Provenance as a first-class output

We instrument the bench so that every measurement emits its own provenance alongside the number. The assay does not just report a value; it reports the conditions under which the value is meaningful.

• Reagent lineage captured at the point of use.
• Instrument calibration state stamped onto each reading.
• Protocol deviations logged as structured events, not lab-notebook prose.

The goal is not more data. It is data that knows what it means.

The payoff compounds: once provenance is structured, an assay becomes a reproducible claim that downstream models can trust without a human re-deriving the context by hand.