Verification & records
A test case and its execution. A checklist and its audit. An instruction and the build of one unit. Each is the same pair: an instrument with ordered prompts, and a record of typed responses about one subject. Avera implements the pair once, as a platform capability.
Instrument
Record
Ordered prompts with expected results and tolerances, authored as items, versioned and approved like everything else.
Executions capture typed responses step by step, citing the exact version of the instrument they ran against; a record can never drift from what was executed.
Campaigns run against named setups; results roll up per instrument and setup, with anomalies raised in place.
CI pipelines deliver results as records citing the exact commit, authenticated with per-job identity and no stored secret. Integrations →
A record is about a subject, and units and lots are items: serials, device history, complaints and vigilance traced to the exact design version each unit was built from.
These aren't modules: they're configurations of the same instrument-and-record engine, which is why they all trace, version, and audit the same way.
Test cases, executions, runs, plans, and anomalies; verification coverage as a live view, not a spreadsheet.
DQ/IQ/OQ/PQ, tools and fixtures, measuring equipment with calibration records and the out-of-tolerance path into nonconformance.
Internal audits as instruments and records: the checklist, the findings, the trail, all items.
Who was trained on what, against which version of the procedure, recorded like every other controlled activity.