Privacy

Last updated 10 September 2026.

This describes what Interview Signal collects, why, and what we cannot see. The security page covers the same ground with more technical detail; where the two differ, the security page is more specific and both are meant to be true.

Who we are

Interview Signal provides a desktop application to recruiting and staffing agencies. For the account data described below we are the data controller. For interview content we are neither controller nor processor, because it never reaches us — see "What we never receive".

What we collect about you as a customer

What we collect about interviews

One row per interview, with exactly these fields:

interview identifier · organisation · seat · role identifier · kit version · source (video, phone or in person) · start time · end time · duration in seconds · paused seconds · consent acknowledgement (notice read aloud, or not required) · jurisdiction label · questions total · questions covered · must-asks total · must-asks covered · follow-ups shown · follow-ups used · scorecard areas · scorecard areas scored · fit score · whether a submittal was exported and in which formats · app version · operating system.

That list is exhaustive. The API refuses any request carrying a field outside it, and the database table has no free-text and no document column, so nothing else can be written to it even by mistake.

We use this for three things and nothing else: to show your team its own coverage report, to make the weekly coaching note, and to see whether the product is failing for someone.

What we never receive

These are written to the recruiter's own computer. We have no access to them, cannot retrieve them for support, and cannot produce them in response to a request, because we do not have them.

Candidates

If you are a candidate who was interviewed by an agency using Interview Signal: we hold nothing that identifies you. Your name, your words and any assessment of you are on the recruiter's computer and are controlled by the agency that interviewed you. Ask them. If you are not sure who to ask, write to [email protected] and we will tell you what we can, which will be that we do not hold your data.

Processing while an interview runs

Short chunks of audio are sent to a speech-to-text provider and text comes back. The chunk is processed and discarded. Text from the transcript is sent to a model provider to produce follow-up suggestions, the scorecard reasoning and the submittal. If the recruiter has attached a CV to the interview, extracts of it go with those requests too, so that a follow-up can probe what the CV claims and the submittal can flag where the CV and the conversation disagree. All providers are used under paid API terms, under which submitted data is not used to train models. Providers are listed on the security page.

Website

How long we keep things

Your rights

Depending on where you are, you can ask for a copy of what we hold, ask us to correct it, ask us to delete it, or object to some processing. Write to [email protected]. We reply within 30 days, usually much sooner, and we do not charge for it.

To delete an entire organisation and every metadata row belonging to it, email us and we will do it within seven days.

Transfers

Our infrastructure and providers are in the United States. Where data is transferred from the UK or the EEA, we rely on the standard contractual clauses with each provider. Bear in mind that the interview content itself is not transferred anywhere: it stays on the machine it was created on.

Changes

If we change what we collect, we will change this page and email account admins before it takes effect. We will not quietly widen the list above — it is the specific promise the product is sold on.

Questions: [email protected].