You're working on a new client brief. A colleague has done something similar before, but you don't know who. The information exists somewhere, but you don't know where. You ask around, search Confluence, scroll through old emails and eventually find three conflicting documents. You decide to work it out yourself. What you discover later: a colleague completed the exact same piece of work two months ago.
This is not an exception. It is the reality at almost every growing organisation.
Knowledge lives in people's heads and personal notes. How do you make it accessible to everyone? Carsten de la Fonteijne, AI Adoption Lead at iO and nominated for the AI100 2027 by Marketing Report, is building the solution: the company brain.
But what exactly is it? And how can you get started today? Read on to find out.
What is a company brain (and what isn't)?
You have probably searched for something in a knowledge base before: on Confluence, SharePoint or an intranet. You search, get a list of results, and then you have to figure out which one is relevant. It is like an endless library where you must hunt down the right book yourself. That takes a lot of time.
If the knowledge base is a library, the company brain is a lightning-fast assistant that reads every book in an instant and gives you the right answer directly. In practice, that means project knowledge, decisions, meeting notes and insights (currently scattered across notebooks, emails and colleagues' heads) become searchable and available to everyone.
A company brain does more than just answer questions. The system learns from every interaction and improves over time. And it does not always wait for you to ask something. When you start a new project or receive a new client brief, the company brain proactively surfaces relevant information. That is how you stop reinventing the wheel.
Case: the company brain in practice
An energy sector service provider runs neighbourhood campaigns to persuade residents to switch to a district heating network. The approach needs to differ per neighbourhood: the demographic make-up determines which resident profiles apply, and each profile calls for a different message and tone of voice. An area dominated by high-value owner-occupied homes requires a very different approach to one with a high proportion of social housing.
The team already has the knowledge. Personas have been developed, strategies are worked out and test campaigns have been ran. But after every campaign round, the findings disappear into a report that not many people read.
Which questions came up most often during the information evening?
Which approach worked in which neighbourhood? which did not?
What do the consultants need to know before they head out?
The answers exist, but they are not available to everyone.
From answers to actions
With a company brain, you can solve this problem. Reports from consultant conversations are automatically processed and analysed. Patterns in residents' questions are identified and shared with the web team, who can address them in FAQs straight away.
The briefing for tomorrow's consultants is also updated automatically, drawing on the most frequently asked questions from today's conversations and the responses that landed well. Each campaign round adds context to the next. Rather than isolated reports, you end up with a learning layer that keeps getting sharper.
How do you keep knowledge up to date?
This is the question that naturally follows. A company brain only works if the information it draws on is accurate and kept up to date. However, nobody opens the intranet every day to log their notes. As a solution, the company brain pulls information from things you are already doing.
You record a meeting
Your personal AI agent transcribes the recording and drafts the notes
You approve the notes
The AI agent shares the notes with the company brain
Your working process does not change. The result simply lands in the right place. On top of that, the company brain keeps itself sharp. If there is duplicate information, or new input contradicts something already on record, it flags it immediately:
'I can see that your notes conflict with a decision that was previously recorded. Which one is correct?'
This keeps the quality of the knowledge intact, even when multiple people are contributing information at the same time.
Now think about what happens when a colleague leaves. Someone with years of project experience takes an enormous amount of contextual knowledge with them when they go. With a company brain, that knowledge does not disappear. It is stored in a structured way and forms the foundation that new colleagues can build on from day one.
So we just gather all data in one system?
No, not everything belongs in a company brain. There are separate systems for contracts, financial data, timesheets and ticket tracking. The company brain does not hold that information directly, but it does have access to those systems and pulls from them only when it is relevant to the question being asked.
That is where the real difference lies. Not a slow, overloaded system stuffed with data, but a smart assistant with access to all the right sources, delivering the right context at the right moment. You can read more about this in our blog on Context Engineering.
How can you start today?
The first step is surprisingly straightforward: start recording all your meetings. Build a working practice that ensures information exists in the first place, then build outwards from there. Ask yourself three questions:
Where does your team lose the most time? Those are the places where a company brain delivers immediate value.
What important knowledge walks out the door when someone leaves? That is what you want to preserve first.
Where does your team keep having to reinvent the wheel? Campaigns, project kick-offs, tender responses — those are the places where a learning context layer pays off the most.
You already have the knowledge
Now you just need the infrastructure to use it. A company brain does not add new knowledge. It makes the knowledge that already exists accessible, findable and usable for the right person, at the right moment.
Organisations that get this right build a foundation on which all further AI applications can develop. For the technical infrastructure, iO Bonzai (opens in new tab) provides the base: a unified AI platform for integrations, governance and AI orchestration. The platform is EU AI Act-aligned and ISO 27001 certified, keeping all data securely within the EU.
Carsten de la Fonteijne
AI Adoption Lead, iO
