This guide concerns Euria, the sovereign artificial intelligence assistant, a full-fledged kSuite service, also available within an app.

Preamble

• The use of AI at Infomaniak is based on a transparent and ethical approach.
• Each AI-integrated service follows strict principles to ensure reliable and respectful use:
  • information protection is guaranteed by full data processing within infrastructures located exclusively in Switzerland, thus ensuring complete control over digital sovereignty,
  • the results generated are never used to train models or for advertising purposes,
  • and AI is not intended to replace humans, but to assist them.

What is Euria?

Designed to offer a European alternative to American tech giants, the sovereign artificial intelligence assistant Euria stands out for its deep integration into the Infomaniak productivity ecosystem, while guaranteeing absolute confidentiality of user data.

Euria's strength lies in its local infrastructure:

• Unlike solutions based on foreign public cloud, Infomaniak's AI is hosted exclusively on its own servers in Switzerland.
• This approach ensures strict compliance with data protection standards and guarantees that no processed information is used to train third-party models or shared outside the secure infrastructure of the host.

Euria uses state-of-the-art language models (LLM) selected for their efficiency and performance, while being optimized to minimize their energy footprint. Infomaniak is committed to total transparency regarding the algorithms used, in line with a technological independence and environmental sustainability approach specific to the company.

Usage limits

⚠ Number of messages per day:

⚠ Web searches per day:

⚠ Number of projects:

⚠ Number of conversations per project:

The transcriptions are unlimited. The image upload is limited to 10 MB. The file upload is limited to 100 MB (except my kSuite/my kSuite+ at 20 MB).

Some Infomaniak services integrate Euria

• kChat
  • The AI Euria integrated into kChat saves time in exchanges: it can rephrase sentences, generate quick responses, correct spelling, or adjust the tone of a message. The user retains full control over the suggestions, which are local to each session only.
• Mail Infomaniak
  • In the Mail app, Euria assists in writing emails. It can generate content from a few keywords, rephrase or structure a message, adapting to different registers. Nothing is sent automatically, and the user can modify the texts at any time.
• Site Creator
  • An AI-powered writing assistant allows Site Creator to quickly generate content tailored to different types of web pages. This service is useful for those launching a site and looking to produce clear, optimized, and relevant texts.
• kDrive
  • On kDrive, artificial intelligence is used at two complementary levels to simplify document management. Euria allows interaction with text files stored in the drive. This module can answer questions about the content of a document, summarize, translate, or extract useful information. It is particularly effective for reading reports, understanding complex texts, or quickly generating an overview of a file.

Responsibilities of the parties

Data provided to the AI

• The AI of Infomaniak is currently based by default on these models.
  • Among them, Mistral, Whisper, DeepSeek, Llama, granite…
• By using a writing assistant, the messages and requests addressed to the AI are exclusively processed in Switzerland by Infomaniak's infrastructure.
• No data is processed or shared with a third party.

Data provided by the AI

• The language model generates text based on the data it has assimilated during its training, devoid of consciousness, intentions, or moral responsibilities.
• It is crucial that the user is aware that the information generated by the model is not always accurate and should not be considered an infallible source of information.
• The user must exercise critical discernment and verify the information provided by the assistant from reliable sources when necessary.
  • For example, if a person uses the model for crucial tasks such as medical, legal, or financial advice, they must take additional precautions to confirm the information from the model.
• In case of inappropriate or dangerous use of this assistant, the responsibility lies with the user who takes these actions.