How does it work?

Models and hosting

Dastra uses pre-trained generative AI models, available from four model families:

Family	Configured models	Hosting
OpenAI (recommended by default)	Fast: gpt-4o-nano · Smart: gpt-5-mini · Large context: gpt-4.1-mini	Azure, France / EEA
Mistral	Fast: mistral-small-2503 · Smart: Mistral-Large-3 · Large context: Mistral-Large-3	Azure, France / EEA
Open source	Fast: Ministral-3B · Smart: DeepSeek-V3.2 · Large context: Llama-4-Maverick-17B-128K	Azure, France / EEA
Custom AI provider	Your own provider via API key (see below)	Depending on your configuration

For the first three families, models are hosted on Microsoft's Azure infrastructure, in France. They are not operated directly by OpenAI or Mistral: Dastra uses Azure managed services, meaning your data does not transit through OpenAI's or Mistral's own infrastructure.

For each family, three model tiers are used depending on task complexity:

• Fast — simple actions: generating descriptions, suggesting tags

• Smart — actions requiring more reasoning: generating structured elements in Dastra

• Large context — actions processing large volumes of data: answering a questionnaire, analysing a processing activity

You can configure the model family in Account Settings > AI Assistant.

Custom AI provider

Dastra allows you to connect your own AI provider via an API key, provided it is compatible with the OpenAI API standard. Supported providers include OpenAI, Google (Gemini), Mistral, Microsoft Foundry, and any compatible self-hosted LLM.

To configure a custom provider, go to Settings > AI Assistant > Custom AI provider, then enter your credentials and map them to the three model tiers (Fast, Smart, Large context).

When using a Custom AI provider, data sent to the model is subject to your provider's privacy policy, not the Azure guarantees described on this page. Review your provider's terms before enabling this option, particularly if your prompts may contain personal data.

What Dastra sends to the model

Dastra only transmits the information necessary for the requested generation: the text entered by the user, and where applicable the structured fields of the object in question (processing activity, AI system, rights request, etc.).

What Dastra does not do

The following points are contractually guaranteed by Microsoft Azure for all models in the OpenAI, Mistral, and Open source families:

• Your prompts and outputs are not available to other Azure customers

• Your data is not transmitted to OpenAI, Mistral, or any other third party

• Your data is not used to train or improve base models

• Models are stateless: no prompt is stored in the model between two requests

Abuse monitoring and human review

Microsoft Azure maintains an automated mechanism for detecting potentially abusive content. In exceptional cases where content is flagged, a sample of prompts may be temporarily retained in a per-customer isolated storage environment for review purposes. For resources deployed within the EEA (which is the case for Dastra), any human reviewers are also located within the EEA.

This retention is exceptional and does not apply to normal platform usage. For more details, consult the Microsoft documentation on Azure Direct Models data privacy.

AI invocation logs

Dastra retains a history of AI assistant calls for the past 90 days, accessible from Settings > AI Assistant > AI invocation logs. For each call, the following information is recorded:

Field	Description
Date	Timestamp of the call
Status	Success or failure of the generation
Operation	Type of prompt used
Model	Name of the model called
Duration	Request processing time
User	Workspace member who triggered the generation

These logs serve as an internal traceability and audit tool. They allow you to identify who used the AI assistant, when, and with which model — particularly useful in the context of a compliance audit or an AI usage review within your organisation.

---

What data is sent to the AI model?

Dastra does not transmit any workspace data to the AI model unless you initiate it. Only the information listed below, depending on the feature used, is included in the request sent to the model.

Feature	Data sent to the model	Potential personal data
Generate a processing activity	Free text entered by the user	No — unless the user mentions individuals in their description
Generate an asset	Free text + URL of the privacy policy provided	No
Generate an actor	Free text + URL or attachment provided	No
Generate a dataset	Free text entered by the user	No
Generate a security measure	Free text entered by the user	No
Generate a privacy notice	Structured fields of the processing activity (purposes, legal basis, retention periods, recipients, rights)	Potentially — if fields contain named references
Generate a description	Free text entered by the user	No
Suggest tags	Text content of the object (name, description)	No
Generate a questionnaire template	Free text entered by the user	No
Answer a questionnaire (PIA, risk analysis…)	Structured fields of the linked processing activity + any custom instructions	Potentially — depending on the processing activity content
Generate a response to a rights request	First and last name of the requester, request message, language, workspace name, purposes, operator name, date and remaining deadline, request status and ID	Yes — first name, last name, and request content
Generate a data breach post-mortem	Structured fields of the breach (description, affected data, measures taken)	Potentially — depending on the content of the breach record
Extract contract metadata	Text content of the transmitted document (attachment or URL)	Potentially — depending on the contract content
Generate a custom document	Instructions entered by the user + any source content	Potentially — depending on the instructions provided
Generate a custom report	Instructions entered by the user	No — unless the user includes data in their instructions
Risk analysis (AI system)	Structured fields of the AI system (description, purposes, data processed, stakeholders)	Potentially — depending on the AI system content
Generate an AI system description	Free text + URL or attachment provided	No
Generate an AI system notice	Structured fields of the AI system concerned	Potentially
Suggest controls / requirements / tests	Context of the object concerned (name, description, framework)	No

Best practice: Avoid including personal data directly in free-text fields (descriptions, custom instructions). Use identifiers or generic terms wherever possible.

Data that is never transmitted

The following are never included in requests sent to the AI model:

• Dastra user credentials and passwords

• Data from other records in your workspace (only the object you are working on is involved)

• Session metadata (name of the logged-in user, IP address, etc.)

• Attachments not explicitly provided within the relevant feature

Retention period for requests

The Azure models used by Dastra are stateless: prompts and outputs are not stored in the model and are not used to retrain or improve base models.

Exception — abuse detection: Microsoft Azure maintains an automated mechanism for monitoring potentially abusive content. If a prompt is flagged, a sample may be temporarily retained in a per-customer isolated storage environment for review purposes. For resources deployed within the European Economic Area (which is the case for Dastra), any human reviewers are also located within the EEA. This retention is exceptional and does not apply to normal platform usage.

For more information, consult the Microsoft documentation on Azure Direct Models data privacy.