TL;DR
Most AI CV tools give you a rewrite and ask you to trust it. We show you every change, which job requirement it addresses, and why we made it. Here is what that audit trail looks like in practice and why we built it this way.
What the audit trail does
When a bullet gets rewritten, you see four things:
- The original text, struck through
- The new version, with the changed words highlighted
- The keyword that was added, as a tag chip
- A one line explanation citing the specific job requirement it addresses
Every change is reversible. Every rewrite is editable inline. Every keyword incorporation is traced back to a line in the job description.
Why this matters for trust
The biggest objection to AI CV rewrites is not output quality. It is trust.
Candidates run their CV through a black-box rewriter, see new output, and cannot defend the result if asked in an interview. They cannot tell whether anything was fabricated. They cannot learn from the changes. So they do not use the output. They go back to rewriting the CV by hand.
When we shipped the audit trail, the share of users who actually export the rewritten CV went up sharply. Support tickets dropped. People started using the tool repeatedly across applications instead of running it once and abandoning it.
The general lesson for AI tools
AI output that people cannot explain back to themselves is worse than no AI output.
If your tool produces something the user cannot defend, justify, or learn from, they will quietly stop using it. Polished output without provenance is a one-shot experience. The audit trail pattern, even in its simplest form, turns that into something users return to.
Top comments (0)