Skip to content

Evee, the AI assistant

On the portal, Evee answers employees' questions from your knowledge base. For agents, it drafts replies, summarizes requests and suggests triage. An admin turns each feature on separately.

For employees on the portal

Employees ask a question in plain language. Evee answers from published articles, cites them, and can check the status of the employee's own requests. When a request is needed, Evee prefills one for the employee to review and submit; it never files a request on its own.

  • A self-service answer is offered before a request is filed
  • Employees see only what they could already open

For agents on a request

On a request's Evee tab, agents can summarize a long thread, get a suggested reply or polish their own draft, and look up similar requests that were already solved. Evee also proposes changes to the status, priority or assignee, or an internal note, and each one waits for the agent to apply it.

  • Every applied change is recorded under the agent's name
  • Draft a knowledge base article from a resolved request
  • Similar requests need semantic search turned on

Triage and reporting

For each new request, triage suggests a priority, request type and team, and sentiment analysis flags frustrated requesters. Managers get a weekly digest that groups the week's requests by theme, and survey comments are summarized the same way.

  • Triage changes nothing until an agent applies the suggestion
  • Describe a queue view in plain English, such as "open hardware requests from last week"

Controls for admins

Each feature stays off until an admin turns it on, and usage is logged by feature, person and team. The cloud plan includes 1 million tokens a month. Admins see a warning at 80 percent, and AI pauses at the limit; there is no overage charge.

  • Confidential requests are never sent to the AI provider unless an admin allows it
  • Customer content is not used to train AI models
  • Self-hosted installs use your own AI provider
  • One setting turns AI off entirely for an offline install

Included in every plan.See pricing →

Questions about Evee AI

Is Evee included in the price?
Yes. The cloud plan includes 1 million AI tokens a month for the workspace, and there is no per-agent charge for AI. Self-hosted installs run on your own provider account.
Which AI model does Evee use?
The cloud edition uses Anthropic, which is listed on our subprocessors page. Self-hosted installs choose from 3 providers: Anthropic, AWS Bedrock, or any OpenAI-compatible endpoint, including models you run yourself.
Can Evee change a request by itself?
No. Evee suggests changes, and an agent decides whether to apply them. On the portal, Evee prefills a request that the employee reviews and submits.
What happens with confidential requests?
They are not sent to the AI provider unless an admin turns on AI for confidential requests, and then only for people who can already open the request. Automatic features such as triage and sentiment always skip them.

Try it with your own data

Create a workspace and set it up with your own teams, request types and settings.