Skip to main content

Call transcript pseudonymisation

Keep your call transcripts longer for coaching and insights, while masking the personal data they contain.

Call transcript pseudonymisation automatically replaces personal identifiers in your call transcripts (names, email addresses, phone numbers) with neutral placeholders after a period you choose. The business content of the conversation stays intact: topics, objections, competitors, company names, job titles.

It works alongside your Data retention policy: instead of choosing between deleting transcripts early or keeping personal data for a long time, you can mask personal data early and keep the transcript for longer.
​
​Talk to your Customer Success Manager to have it activated on your account.

Key benefits

  • Data minimisation under GDPR: personal data is only kept in identifiable form for as long as you need it.

  • Long-term insights: keep transcripts long enough for year-over-year analysis, coaching and competitive intelligence.

  • Business context preserved: only personal identifiers are masked, so your teams keep the value of the conversation.

  • Historical data covered: the same processing applies to transcripts already stored in Modjo.

💡 Pseudonymisation vs anonymisation: pseudonymised data can no longer be attributed to a person without additional information. It reduces privacy risk but is not full anonymisation. We recommend validating your retention and pseudonymisation periods with your DPO or legal team.

1. What gets pseudonymised

Modjo detects the following personal identifiers in the transcript and replaces them with placeholders:

Personal data

Replaced with

Email address

[email]

Phone number

[phone number]

Name of an internal user (your company)

[internal user first name], [internal user last name], [internal user full name]

Name of an external participant

[external user first name], [external user last name], [external user full name]

How names are detected?

Modjo uses the people it already knows for this call, meaning the call participants and the contacts of the account the call is linked to. A name is replaced even when only the first or last name is said.

Example

Original: "I accepted the bot because Mariane will want the call summary. Have you spoken with Louise before?"

Pseudonymised: "I accepted the bot because [external user first name] will want the call summary. Have you spoken with Louise before?"

Here, Mariane is a known contact on the account, so her name is replaced. Louise is not a participant or a known contact, so her name is kept.

🚨 Limits to be aware of

  • Names of people who are neither call participants nor known contacts of the account are not replaced.

  • Names that were transcribed with a spelling error may not be recognised.

  • Postal addresses are not pseudonymised yet.

  • The call recording is not modified. To remove personal data from audio, set a call recording deletion period equal to or shorter than your pseudonymisation period.

2. How it works

Pseudonymisation runs once per call, at the end of the period you set. Here is the lifecycle of a call transcript:

  1. Call import: Modjo generates the full transcript, then the AI features (AI summary, CRM filling, AI call scoring…).

  2. Pseudonymisation period reached (e.g. 6 months): Modjo generates the pseudonymised version of the transcript, then permanently deletes the original. From then on, everyone sees the pseudonymised transcript.

  3. Transcript deletion period reached (e.g. 12 months, if enabled): the pseudonymised transcript is deleted, as described in Data retention policy.

🚨 Warning:

  • Once a transcript is pseudonymised, the original version is permanently deleted and cannot be recovered.

  • Pseudonymised transcripts are not yet available on the call page. They can still be retrieved through API.

3. Configuring pseudonymisation

Admins can enable pseudonymisation in Security Settings, in the same place as the Data retention policy.

A. Access settings: go to Security Settings > Data Retention Policy, then click Edit.

B. Enable pseudonymisation: in the Period before call pseudonymisation section, turn on Pseudonymise call transcript.

C. Choose the period: set how long after call import before transcripts are pseudonymised (e.g. 6 months).

D. Check your other retention periods:

  • Call transcript: if you also delete transcripts, set a deletion period longer than the pseudonymisation period (e.g. pseudonymise at 6 months, delete at 12 months).

  • Call recording: the audio is not pseudonymised. We recommend deleting recordings at or before the pseudonymisation date.

E. Save your configuration: click Save at the top of the page.

💡 Existing calls: once saved, the setting also applies to transcripts already stored in Modjo. Any call older than the period you set will be pseudonymised.

4. Using Modjo once a transcript is pseudonymised

The pseudonymised transcript replaces the original everywhere in Modjo. Features that read the transcript keep working, but show placeholders instead of personal identifiers.

What is kept as is

  • Call details: Modjo users on the call and CRM associations (contact, deal, account).

  • AI summary: generated before pseudonymisation, it stays visible.

  • AI features keep working on the pseudonymised transcript.

  • Call recording: kept until its own deletion period, if one is set.

Email data is not affected by this feature.


Related articles

Did this answer your question?