Simplycure integrates artificial intelligence tools to support you day to day. This page explains which data is involved, how it is protected, and why you can use it with confidence.
In short: all data processed by the AI stays on infrastructure certified as a Health Data Host (HDS), hosted in Europe. It is never used to train models, nor passed on to third parties.
What the AI does in Simplycure
Two features use artificial intelligence in your practitioner space:
Health questionnaire analysis: an automatic summary of your patient's answers is generated for you, with the key points to bear in mind.
Conversational assistant: you can ask questions to the assistant, which can consult your catalogue, your patient records and your recommendations to answer you.
Good to know: a fourth AI feature allows questionnaires to be imported (PDF, Word). It only processes blank form templates: no patient data is involved.
Which infrastructure? Which certification?
All AI processing runs on Amazon Web Services (AWS), in the Europe (Ireland) region. AWS is certified as a Health Data Host (HDS) by the Agence du Numérique en Santé (ANS).
The service used to run the language models, Amazon Bedrock, is included within the scope of this certification.
Useful info: you can view the AWS HDS certificate on the official ANS website: esante.gouv.fr > Labels & certifications > HDS.
Which AI model is used?
Simplycure uses a language model developed by Mistral AI, deployed via Amazon Bedrock. In practice, this means:
The model runs entirely within the AWS infrastructure.
The data never leaves this environment.
Mistral AI, as a company, receives no customer content: only anonymised and aggregated usage statistics are shared with it, in accordance with the terms of use of its service on Amazon Bedrock.
Which data is sent to the AI?
Depending on the feature used, here is what passes through the model:
Questionnaire analysis
All of the patient's answers are sent to generate the summary. If the questionnaire includes identifying data (first name, last name, date of birth), this also passes through, but always within the HDS-certified AWS infrastructure.
Chat about a questionnaire
The questionnaire answers serve as context for the conversation. The messages you send are also transmitted and may contain health information.
General assistant
The assistant can consult your catalogue, your patient records and your recommendations. The data returned (which may include patient names and contact details) passes through the model as part of the conversation.
Point of attention: the general assistant may have access to your patients' data in order to answer you. If you would like to know more about the exact scope of the data that is accessible, do not hesitate to contact us.
What concrete guarantees?
On-demand processing: data is only sent at the moment of processing. It is not kept afterwards.
No training: your data is never used to improve or train the AI models.
Data in Europe: all processing takes place in the European Economic Area (Ireland).
Full encryption: TLS 1.2+ in transit, AES-256 at rest. No health data in the logs.
Secure access: the AI features are reserved for authenticated users. Logical separation between customers.
HDS certified: infrastructure certified as a Health Data Host by the ANS.
Shared responsibility
AWS ensures the security of the cloud infrastructure (security of the cloud). Simplycure ensures the security of the application and the data within this infrastructure (security in the cloud), in accordance with the AWS shared responsibility model.
Frequently asked questions
Can my health data leave Europe?
No. All processing takes place within the AWS infrastructure in Ireland, in the European Economic Area. The data never leaves this environment.
Does Mistral AI have access to my patients' data?
No. Mistral AI, as a company, receives no customer content. Only anonymised and aggregated usage statistics are shared with it, in accordance with the terms of use of its service on Amazon Bedrock.
Can I use these features fully lawfully in my practice?
The infrastructure is HDS certified in accordance with the ANS requirements. If you have specific questions about your situation or your organisation, we recommend consulting a data protection officer (DPO).
Is my data used to improve the AI?
No. Your data is never used to train, fine-tune or improve the AI models. Processing is strictly limited to the duration of each request.
A question about the security of your data? Our team is here to answer you: write to us at [email protected] or via WhatsApp on +32 470 18 14 60.