Transcription
Transparency is central to us at Intric. When you use the Transcription feature to process and summarize recorded audio files, specific processes are in place to ensure your privacy and secure data handling.
The process from your audio upload to a finished transcription and summary occurs through a secure, SSL/TLS-encrypted interaction between the Intric platform (where your data is managed) and the LLM provider you have selected (e.g. Berget AI).
Step-by-step: How your data is handled
Section titled “Step-by-step: How your data is handled”All transfers between Intric and its sub-processors occur over secure, encrypted connections.
Step 1 — User uploads an audio file
Section titled “Step 1 — User uploads an audio file”The user uploads a recorded audio file to Intric in the browser, or uses the transcription feature.
Data sent to Intric’s server:
- The audio file
- File name and metadata
All content is sent over an encrypted connection from your browser to Intric’s servers.
Step 2 — Intric receives, encrypts, and stores the file
Section titled “Step 2 — Intric receives, encrypts, and stores the file”Intric’s server receives the audio file, encrypts it, and stores it securely in Intric’s database, which is operated by a Swedish sub-processor.
Before the audio is sent to the AI model, Intric removes identifying metadata. No information about who uploaded the file or which organization it belongs to is attached.
Step 3 — Intric calls the assistant’s selected transcription model
Section titled “Step 3 — Intric calls the assistant’s selected transcription model”In this step, relevant data from Intric’s servers is sent to the assistant’s selected transcription model.
Data sent from Intric’s server:
- The audio file
- System instructions for transcription and summarization
The transfer between Intric and the transcription model occurs over an encrypted connection.
What happens at the provider: The model generates a text transcription and then a summary. The provider has zero context about the user’s or organization’s identity.
Step 4 — Response to Intric
Section titled “Step 4 — Response to Intric”The transcription model’s response is sent to Intric’s server, which receives and stores the information encrypted in its database.
Data sent from the transcription model to Intric:
- The transcription model’s generated result (transcription and summary)
Immediately after the result is sent back to Intric, both the user’s audio input and the generated result are deleted from the transcription model’s server.
Step 5 — User sees the response in the browser
Section titled “Step 5 — User sees the response in the browser”The response is displayed to the user in the browser.
Data stored on Intric’s servers:
- The generated response (transcription and summary) and the history from the user’s interaction with the assistant
- The uploaded audio file
- Metadata about who sent the request
Data sharing and privacy
Section titled “Data sharing and privacy”To protect your and your organization’s privacy, we apply the principle of data minimization. This means the sub-processor only gets access to the content absolutely necessary to perform the task — no user identity ever leaves your infrastructure.
We have strict zero data retention clauses in all our contracts with language model sub-processors. This guarantees that your audio files and the generated texts are never saved by the provider after the response is returned, nor is the information used to train their AI models.
In the table below, you can see exactly what data is sent to the sub-processor and what does not leave Intric’s servers.
| Sent to the transcription model | Not sent to the transcription model |
|---|---|
|
|
Data retention and deletion
Section titled “Data retention and deletion”Transcription history follows the same principles as other assistant interactions — depending on whether the assistant is a personal assistant or in Spaces, the Creator or Admin controls deletion.
Metadata about how users use features in Intric can be stored for longer periods and is available to administrators.