Severa AI is an AI-based service designed to assist users by answering questions about their own Severa environment and data, as well as executing requested commands directly in the Severa Beta interface. Questions or commands may relate to work hour entries or projects. 

Severa AI is currently in the pilot phase and is available to users within the scope of the Severa Beta interface redesign. Severa AI responds in the user's language, either Finnish, Swedish, or English.

At the beginning of the pilot phase, not all data has been made available to Severa AI at once. We are currently focusing on the most essential information, and the selection will expand based on user feedback.

Is your company interested in trying the Severa Beta interface? Welcome! Contact Severa customer service to start the trial: severa.support@visma.com 

How does Severa AI work?

You can ask Severa AI questions related to the available data and request it to perform, for example, time entries for desired projects on your behalf. Severa AI has access to the same information that you can access in the interface based on your permissions.

Tips for boosting everyday work with Severa AI

☑️ Ask Severa AI to log hours for you: Based on your request, Severa AI suggests entries, and you can accept them directly or fine-tune them before accepting.
☑️ Summaries of work hour entries or absences: Quickly compile all hours you have logged for a project with the help of Severa AI. Just specify whether you want, for example, data for a specific time period or specific phases.
☑️ Reporting of billable hours: You can retrieve a summary of billable hours, for example by customer, project, or phase. This helps you see how much billable work has accumulated for different targets.

Tips in practice

Chatting with AI is easy when you remember a few basic rules. The more precisely you can describe your need, the better the answer you will get.

✅ Do's (do this):

  • Be specific: Clearly define what you are looking for (e.g. a specific time period, project, or person).
  • Make use of the conversation: The AI remembers previous messages in the same chat. You can continue by asking: "What about the previous week?" or "Can you group these by customer?"
  • Provide context: Tell the AI exactly which project or whose data you are referring to.
  • Stay on topic: To keep answers as accurate as possible, we recommend starting a new conversation whenever you change the subject.

❌ Don'ts (avoid these):

  • Do not ask too much at once: Avoid complex, multi-part questions in a single message. Split large requests into smaller steps.
  • Do not forget to verify information: Although the AI is smart, it may sometimes make mistakes. Always verify critical figures directly in the system.
  • Avoid overly general questions: "How are projects progressing?" is too broad. Instead, ask more specifically.
  • Do not share sensitive information in the conversation: Do not share sensitive personal data with the AI, such as social security numbers, online banking credentials, or passwords.


What is a good prompt? 

A prompt is an instruction, question, or task given to the AI (in this case, Severa AI). A good prompt is a clear command or question from which the AI immediately understands your goal.

  • Weak prompt: "How many hours are unlogged?" In this prompt, the AI does not know whose hours, from which project, or for what time period the information should be retrieved.
  • Good prompt: "List all my unlogged hours for this week in project X." In this prompt, it is clearly stated to the AI what should be retrieved and on what basis.

Data sources

Severa AI utilizes multiple data sources to provide accurate and useful answers. Severa AI reads data from selected sources that are specifically related to work planning and tracking. Available data includes, among others:

  • Time entries: Work hour entries and absences
  • Project management: Projects, their phases, and related activities
  • Advanced resourcing: Information on how cases have been planned and allocated
  • User information: Basic information about users in the system

Security and privacy

By using the Severa AI function, the user agrees that the AI processes their data in Severa. The AI may occasionally produce incorrect or inaccurate information. The user should always verify critical information and results themselves in Severa before utilizing them in decision-making.

  • Case history: Conversations are stored in the user's own case history, where the user can also return to previous conversations. The case history is retained for six (6) months from the latest message, after which inactive conversations are automatically deleted.
  • AI training & development: Severa AI does not collect or use conversations as automatic training material. However, functionality is improved by reviewing previous conversations manually to ensure quality.
  • Permissions: Severa AI can use information as a basis for its answers and commands only from projects to which the user has permission to access in the interface. This ensures that Severa AI does not access any information to which the user has not been granted permission in the interface. Read more about Severa's permission levels: Käyttöoikeudet tarkemmin
  • EU AI Act: Severa AI has been developed and audited to meet the requirements set by the EU AI Act. You can read more about the EU's regulatory framework for AI on the European Commission website.
  • Feedback & development: We welcome feedback and experiences from using Severa AI so that we can further develop the service.


Keywords: Severa, AI, Severa AI, artificial intelligence



Did you find it helpful? Yes No

Send feedback
Sorry we couldn't be helpful. Help us improve this article with your feedback.