Created with AI.
Use this guide for a bounded trial in a familiar workflow. Aim for better work and documented learning, with clear roles for leaders, employees and technical owners.
Find a recurring task whose quality the team understands. Establish current effort and errors. Prioritise a verifiable trial instead of requiring AI for every task.
Describe the problem and expected improvement without mentioning features. Then assess whether Copilot or Cowork fits.
Agree permitted tools, data and actions with responsible organisational owners. Name a process owner, reviewer and error contact. Provide an easy route for help.
Separate drafts from actions that change commitments, send information or affect customers. Define a stop rule when boundaries or quality cannot be maintained.
Have colleagues try the same fictional task against clear criteria. Discuss benefits, gaps and necessary checks. Use the power user for support rather than making them the only capable reviewer.
Allow time for practice and review. Collect useful instructions with their purposes and limitations, without customer data.
Compare similar tasks with and without assistance, including preparation, verification and rework. Check equivalent quality. Separate measured effects from reported perceptions.
Practice case: Previously 50 minutes; now ten preparation, ten generation and 20 checking. The saving is ten minutes. If corrections add 15 minutes, total effort is five minutes higher than before.
Continue, change or end the method based on documented results. Assign ownership, training, follow-up and reassessment when sources or processes change.
Write a brief decision covering benefit, quality, remaining risk, ownership and review date. Login counts alone do not justify expansion.
The working routines and fictional examples above are ProsessPilotene learning material. Use the official product documentation below for product context; availability depends on your organisation's setup.
Ask a question or share what helped you.
Artificial intelligence | Guide
The instructions determine whether the agent is useful or embarrassing. This guide shows what actually works.
Artificial intelligence | Guide
Publishing is the final and most visible step. This guide shows the sequence that ensures a controlled launch rather than one full of surprises.
Artificial intelligence | Guide
A first agent should answer a few questions well rather than many poorly. This guide walks you through the setup and the choices that determine whether customers get the help they need.
Artificial intelligence | Guide
The knowledge source determines what the agent can answer. This guide shows you how to connect it without exposing anything that was not intended for customers.
Customer Service | Guide
This guide draws on the earlier technical article and demonstration. Adapt the steps to your organisation's configuration and responsibilities. The original demonstration transcript is available in the Norwegian version of this guide.
Artificial intelligence | Guide
Use this routine for one writing, meeting or analysis task. Give a clear brief and retain control of the output in Microsoft 365 Copilot or Copilot Cowork where available. Continue in the business-area learning path

Automation | Article
OpenAI’s Data agent could investigate company data and build shareable interactive dashboards. Access controls and source trails needed to follow analysis and action through the workflow.

Customer Service | Article
WEM MCP tools enabled time-off requests, shift swaps and shift bids in Service Agent. Administrators could enable each tool for a role, team or pilot group.

Information security | Article
On 4 August 2026, the Norwegian government announced that its proposed AI Act would return to consultation after changes in the EU simplification package.

Information security | Article
The EDPB published new guidelines on web scraping for generative AI and adopted final blockchain guidelines.

Automation | Article
Copilot administrators gained central controls for consumption-based cost. Purview controls extended to Cowork, and agent metrics could connect with organisational context.

Customer Service | Article
Service Agent became generally available across Dynamics 365 and Microsoft 365. Workforce Engagement Management connected demand, staffing and quality work.
Do you have a question or experience to share?
Be the first to contribute.