AI needs more than a good prompt.
Company knowledge does not live in one place. It sits in documents, customer context, conversations and people’s experience. For GoodUp, the question started with the basics: how do we give AI the context to contribute?
The ambition was a shared company brain. Knowledge about the organisation, its ways of working and its relationships. With clear agreements on what should and should not be shared.
Understand first. Then connect.
We started with context: what does an AI colleague need to know about GoodUp? That led to a structure for company information, customer records and working agreements.
We then worked in focused steps. Map the sources. Set up an initial knowledge layer. Test what belongs there in practice. Workshops connected the technical build to the team’s daily work.
- 01
Organise
Give company context, sources and working agreements a clear place.
- 02
Connect
Connect existing information to a shared knowledge layer.
- 03
Test
Use focused pilots to check what works and what still needs attention.
From insight to shared knowledge.
A company brain needs to grow with the business. So we also built a route for bringing insights from AI work sessions back into shared context.
New content gets a destination and is checked. Where a review is needed, a private Slack review helps people decide what can move forward. Approved content can then enter the shared knowledge layer.
Knowledge that works with you.
And grows with you.
From what your company knows to what your team does — and back again.
Select a step to explore how it works.- Company & customers
- Documents & ways of working
- Conversations & work sessions
Who we are · who we help
How we work · what we learn
Working with relevant context
- Prepare customer conversations
- Draft proposals
- Find company knowledge
New insights proposal review shared context
The return path always goes through review, never straight into the brain.Start with what your company already knows.
Company information, customer context, documents and working agreements form the foundation. You choose which sources matter; connecting a source does not mean copying everything.
The team builds with us.
Alongside the system, there was space to work with AI together. The workshops brought the team’s questions, processes and potential applications to the table.
This connects Company OS to the people who will use it: technical development, practical training and working agreements belong together.
A foundation. And a next step.
The work delivered a shared knowledge structure and a functioning route for controlled publication. In the pilot, session insights were published while other content was held back because of its sensitivity.
That is a foundation to build on. Further source connections, document use and wider team adoption need their own tests. We distinguish what has been proven from what comes next.
- Shared context
- Company knowledge, customer context and working agreements in a clear structure.
- Controlled publication
- A route from new insight through review to shared knowledge.
- Learning through practice
- Focused pilots and workshops as a foundation for further adoption.