Signatures & data integrity

DASH is built so that the data behind every certificate is trustworthy: attributed to a real person, recorded at the time it happened, and impossible to alter once signed. This page explains what a signature captures, why signed records are locked, and how DASH supports the ALCOA+ principles that regulators expect.

What a signature captures

When you sign a certificate, DASH binds your identity and the moment of signing to that specific record:

Who signed — your name is recorded against the certificate.
When — the exact date and time of signing.
What data — the signature is bound to the certificate’s content, not stored loosely alongside it.

The signing name you type must match your profile name exactly, so a signature can only ever be applied by the account it belongs to. The event is also written to the audit log, giving you a second, independent record of the signature.

Why signed records cannot be changed

A signature seals the certificate. After signing, the record is locked — its data cannot be edited. The signature is bound to the exact content that was signed, so any change would break that binding and be immediately detectable. This is what “tamper-evident” means: you do not have to trust that nobody altered a signed record, because an alteration cannot pass unnoticed.

If signed data turns out to be wrong, you do not edit it. Instead you re-process the certificate or create a new one, and the original signature and its audit trail are always retained. See Sign a certificate for the full flow.

How DASH supports ALCOA+

ALCOA+ is the set of data integrity principles used across regulated pharma. Here is what each one means in plain terms and how DASH supports it:

Attributable — every action is tied to a named user. Certificate creation, edits, signing and channel submissions all record who did them, backed by the audit log and transaction logs.
Legible — data is stored in open, readable formats and stays readable over time; extracted values keep both their original label and a normalised version.
Contemporaneous — timestamps are set by the system at the moment each action happens, not typed in after the fact.
Original — the first capture of the data is preserved. Your corrections are applied as a non-destructive overlay, so the original ingested or extracted data is never overwritten.
Accurate — validation rules and, for AI extraction, confidence indicators help surface fields that need a closer look, and only authorised people can sign. See Validation rules.
Complete — nothing is quietly dropped. All fields are recorded, including empty ones, and every result row is captured rather than summarised.
Consistent — records are ordered chronologically and status changes follow a defined lifecycle, so the sequence of events is coherent.
Enduring — records are stored durably and retained, rather than living only in transient or editable form.
Available — records stay accessible to authorised users through the app for review and inspection.