Skip to content
BURNABY · L-1B FIELD GUIDE

What technical records make useful evidence?

Sources checked:

THE DIRECT ANSWER

Internal design documents, architecture decision records, runbooks and implementation guides naming the employee as author or maintainer, significant incident write-ups, and code ownership records where policy allows sharing them. Add the technical lead's written description of the work that routes to this person and why.

Choose records a non-engineer can read

Reviewers are not evaluating engineering quality; they are checking whether the described knowledge is real and company-specific. A short authored guide with the employee's name and a date usually communicates more than a large export from a source control system. Where a document is too technical to stand alone, pair it with two or three sentences of context.

Check what may leave internal systems before assembling anything. Add one step before anything is gathered: agree what may lawfully and contractually leave the company's systems. Customer names, licensed third-party material, security architecture and anything covered by a confidentiality obligation each need a decision, and the decision is far quicker when made once at the start than repeatedly under deadline.

Where material cannot be shared, record that it exists, who holds it, and why it is restricted, so a reviewer encounters a governed choice rather than an unexplained gap. Then keep the approver's decision itself in writing. Hypothetical example: a telemetry pipeline specialist's most persuasive evidence is an internal design document naming three customers, and rather than abandoning it the team produces a redacted version with a covering note from the security lead explaining what was removed and on what basis, which preserves the substance without creating a problem elsewhere.