Choose modules
Open Manage organizations and the Modules tab for the organization you want.
- Check which modules are available on your plan.
- Enable the modules you need using their toggles.
- Then check the navigation. Some features live as sub-tabs inside another module.
- If a member doesn't see an enabled module, also check their access role.
Set up the weekly summary
The AI weekly summary rounds up things like tasks, sprint progress, time tracking, budgets, metrics, and milestones, and posts the result to chat.
- Open the AI tab and the weekly summary section.
- Enable the feature and set the day of the week, time, and time zone.
- Check the target channel and its available settings.
- Read and confirm the required consent the first time you enable it.
- Save, and check the result in chat after the scheduled run.
Set up weekly KB synthesis
The separate weekly KB synthesis (beta) requires Chat and Knowledgebase. It turns weekly reviews and project context into a draft covering Success Patterns, Risk Patterns, Decision Log, and Lessons Learned. Set the schedule, time zone, and consent here too. The draft appears in the admin chat; a Knowledgebase article is only created once you review and adopt it.
Check results and change settings
Check AI results for errors and unverified conclusions. You can change the schedule later, or disable the feature and save. Messages already generated and manually documented insights remain in place.
Frequently asked questions
Does enabling a module automatically grant every member access?
The module selection is the foundation for the organization. Custom access roles can further restrict individual people's access.
Are the weekly summary and weekly KB synthesis the same thing?
No. The summary delivers a recap in chat. The synthesis creates a structured knowledge draft from weekly reviews for the admin channel.
Does this permanently train the AI automatically?
The synthesis derives possible insights from existing data. Reviewed, later-usable knowledge comes from the documentation itself; this doesn't promise any training of the underlying AI model.