How, where, and why Akompa processes your data
Last updated: August 6, 2026
The registry below presents the main types of data processed by Akompa, the reasons they are processed, the technology service providers participating in their processing, the applicable security measures, and the territories involved. It is intended, among other things, to help professionals and organizations using Akompa Notes complete, when necessary, their privacy impact assessment.
| Type of data | Why this data is processed | Providers involved | Security | AI models | Transit outside Quebec | Long-term storage outside Quebec | Long-term storage in Quebec |
|---|---|---|---|---|---|---|---|
Audio recordings | To generate transcripts, which are then used to generate notes | storage: Amazon Web Services (Quebec), transcript generation: Modal Labs | ISO 27001, SOC 2, AES-256 encryption at rest and in transit (TLS), data processing agreement (DPA) | Akompa-controlled open-source model. Data is not used to train the model. | Yes, United States | No | Yes, max 24h |
Transcripts, initial notes in processing, and note examples of a custom template | To generate notes and custom note templates | orchestration: Modal Labs, AI: OpenAI | ISO 27001, SOC 2, AES-256 encryption at rest and in transit (TLS), DPA | Various OpenAI models. Data is not used to train the models. | Yes, United States | No | No |
Client identifiers (name, ID, or alias), revised notes, and transcripts | To allow you to access notes and transcripts from past sessions and organize them by client | Amazon Web Services (Quebec), Supabase (Quebec) | ISO 27001, SOC 2, AES-256 encryption at rest and in transit (TLS), DPA | none | No | No | Yes, retention depends on configuration (default 24–48 hours) |
| Type of data | Why this data is processed | Providers involved | Security | AI models | Transit outside Quebec | Long-term storage outside Quebec | Long-term storage in Quebec |
|---|---|---|---|---|---|---|---|
Name, email address, and other waitlist information | Signup request management, user profile creation, secure authentication, and technical support | Amazon Web Services (Quebec), Supabase (Quebec) | ISO 27001, SOC 2, AES-256 encryption at rest and in transit (TLS), DPA | none | No | No | Yes, for the lifetime of the account |
Financial information | Credit card payments | Stripe | PCI-DSS Level 1, SOC 1 & 2, AES-256 at rest, TLS 1.2, DPA | none | Yes, United States | For the term of the agreement and any period required by law (e.g., financial record keeping). | No |
First name and email address | Sending emails required for the web application to function | Resend | SOC 2 Type II, AES-256 at rest, TLS in transit, DPA | none | Yes, United States | Yes | No |
Name, email address, and other waitlist information | Sending optional emails to users (e.g., reminders or resource sharing), with unsubscribe available from the link at the bottom of each message | MailerLite | ISO 27001:2022, AES-256 at rest, TLS in transit, DPA | none | Yes, European Union | Yes | No |
Technical metadata (browser, operating system, etc.) and masked recording of your navigation | Diagnosing technical errors and improving application stability | Sentry | SOC 2 Type 2, ISO 27001, AES-256 at rest, TLS in transit, DPA | none | Yes, United States | Yes | No |
Pages viewed, actions taken in the application, and masked recording of your navigation | Understanding how the application is used in order to improve it, and reproducing reported technical problems | PostHog (European Union region) | SOC 2 Type 2, TLS in transit, DPA | OpenAI and Anthropic models, via PostHog's AI-assisted analysis features. Data is not used to train the models. | Yes, European Union | Yes | No |
Name and email address, and video conversations with our team and their transcript | Receiving and sending emails (e.g., about your questions and comments), and videoconference discussions about Akompa Notes | Google Workspace | SOC 1/2/3, ISO 27001, ISO 27017, ISO 27018, AES-256 at rest, TLS in transit, DPA | Gemini, for meeting transcription and summaries and for searching emails. Data is not used to train the models. | Yes, United States | Yes | No |