Knowledge sharing platform
How knowledge gets from one team to another when nobody publishes it
Inside a team, knowledge is shared all day — in threads, meetings and chats. The problem is everyone outside it, who cannot see those conversations and does not know they happened. SecondBrain 365, an AI agent inside your Microsoft 365, turns those conversations into pages the other teams can read.
What a knowledge sharing platform is for
A knowledge sharing platform is software that lets what one person or team knows reach the colleagues who need it — decisions, procedures, answers, context — instead of staying in an inbox, a meeting or somebody's head.
The short answer to how knowledge crosses between teams: by one of three routes. Someone asks a person who knows; someone publishes it for everyone; or it is written down where the work happens, as a by-product. Most platforms are built for the second route. Most knowledge actually travels by the first. This page is about why, and about what the third one needs to work.
Why knowledge stays inside the team
It is not hoarded. It is shared with the people who were already there, and stops at the edge of the group.
49%
of developers often find themselves answering questions they have already answered before. Inside the team, the same knowledge travels again and again, one person at a time.
Stack Overflow Developer Survey 2024 — about 28,700 respondents
45%
say knowledge silos prevent them from getting ideas across the organisation. Outside the team, the same knowledge does not travel at all.
Same survey, same respondents
The walls also get thicker with distance. When Microsoft studied the email, calendars, chats and calls of 61,182 of its own US employees in the first half of 2020, it found that firm-wide remote work made the collaboration network more static and siloed, with fewer bridges between disparate parts. The authors' conclusion: that makes it harder for employees to acquire and share new information across the network.
Sources: Stack Overflow, 2024 Developer Survey (the sample is software developers, not all employees); Yang et al., Nature Human Behaviour, 2021.
Three routes between teams, and where each one stalls
Every organisation uses all three. The question is which one you are counting on.
Someone asks someone
“Who knows how the EU migration is going?” — two forwards later, the right person answers.
Works
For anything complex. A conversation carries judgement and context that no page does.
Stalls
When you do not know who to ask, or that there is anything to ask about. And the person who knows answers the same question for the fifth time.
Someone publishes it
After the decision, somebody writes it up on the intranet for everyone else.
Works
Once it exists. Anyone can find it, at any hour, without interrupting anyone.
Stalls
At the writing. It is the same knowledge a second time, for readers the author has never met, and it is the first task dropped in a busy week.
Written down as the work happens
The decision is recorded from the meeting where it was taken, the day it was taken.
Works
Nobody writes anything twice, and the record exists before anyone outside the team needs it.
Stalls
Someone has to decide which conversations count. And it records what was decided, not how to do the hard part — for that you still need the person.
Most knowledge sharing tools are a better place to publish. That helps with the second route only after somebody has done the writing — which is exactly the step that stalls. If your teams already share through a wiki and you are weighing a move inside Microsoft 365, that is a separate question: replacing Confluence as your sharing space.
What a page can carry to another team, and what it cannot
Writing things down solves the finding. It does not solve all of the transfer.
A page carries
What was decided, when, by whom, and in what words. Enough for someone in another team to act on it — or to know that there is something to ask, and who to ask.
A page does not carry
The judgement behind a hard call, or how to do something complex. That still moves person to person, through people who work closely enough to explain it.
Research on product teams found the same split. In a study of 120 product-development projects across 41 divisions of a large electronics company, loose contacts in other units helped teams find useful knowledge quickly, but slowed them down when the knowledge was complex: transferring it needed a close working relationship (Hansen, Administrative Science Quarterly, 1999). A cross-team page works like one of those loose contacts. Its job is to make the knowledge findable and to point to the person — not to replace the conversation.
Making one team's knowledge readable by the others
Four changes that need no new software. Each one lowers the wall an outsider hits.
- Instead ofOpening the team's channel to the whole company
- TryKeep the discussion where it is and publish the outcome: what was decided, when, and by whom. Other teams need the result, not the forty messages before it.
- Instead ofWriting for the people who were in the room
- TrySpell out code names, acronyms and team shorthand the first time they appear. It is the first wall an outsider hits, and the author no longer sees it.
- Instead ofAn anonymous page
- TryEnd with a name. For anything complex, the most useful thing a page can do is tell the reader who to talk to.
- Instead ofA new portal that people have to remember to visit
- TryPut the page where people already search — the intranet, SharePoint, the assistant they already ask — with permissions that let the other teams read it.
Pages like these usually end up in one place — where shared knowledge ends up: a company wiki. Once other teams rely on them, the next problem is keeping those pages true once other teams depend on them.
Where SecondBrain 365 fits
The third route, without anyone writing twice
Inside the Platform team
Email · Marco Bianchi · 2 AprFrankfurt is ready. We can move the EU tenants to FRA2 after the April freeze.
Platform weekly · 3 Apr · transcribedAgreed: EU customers move on 12 May. APAC stays where it is.
- Team chat · 4 Apr
Heads-up: the cutover needs a two-hour window.
The conversations stay in the team
Platform KB › Data residency
EU customer data moves to FRA2
- What
- EU customer data moves to the Frankfurt region, FRA2
- When
- 12 May, two-hour maintenance window
- Decided
- Platform weekly, 3 Apr
- Ask
- Marco Bianchi
Sales — another team
When do our EU customers move to the new region?
Answered from the page
On 12 May, to Frankfurt (FRA2). Decided by the Platform team on 3 Apr — Marco Bianchi knows the details.
Only if Sales is allowed to read the page
SecondBrain 365 is an agent with its own identity and company email inside your Microsoft 365. A team copies it on an email, invites it to a Teams meeting, adds it to a Teams group or shares a document with it. From what it is shown, it writes pages in a knowledge base on the team's SharePoint site, links related pages, and highlights the people, customers and acronyms they mention.
Only the members of the team's workspace can feed it. The conversations stay where they were; what comes out is the page. You can require approval before anything is written, and every page stays editable.
Who can read the pages is ordinary SharePoint permission, set the way you already set it. The pages can also be a source for Microsoft 365 Copilot, which only shows people what they are allowed to see (Microsoft's documentation) — so a colleague in another team can get the answer without knowing the page exists, if they have access to it.
It only knows what it is shown: a decision taken in a corridor crosses no boundary until someone puts it in writing where the agent can read it. The mechanism is on how SecondBrain 365 turns a conversation into a page; a typical case is product decisions that sales and support can read.
Microsoft's own attempt at bringing one team's knowledge to everyone else was Viva Topics, retired in 2025 — what it left behind in your tenant, and what fills the gap.
FAQ
Questions about sharing knowledge between teams
Mostly emphasis. A knowledge base is the collection of pages that answer questions; a knowledge sharing platform is the whole way knowledge moves from the people who have it to the people who need it, which usually includes a knowledge base. The distinction matters because a knowledge base only shares what someone has put into it.
Not by opening every team's conversations to everyone — that buries the useful part. Publish the outcomes (decisions, dates, owners) where other teams already search, write them so an outsider can read them, name the person to ask, and make recording them part of the work rather than a task for later.
Make it cost less than not sharing. People already explain things constantly, to whoever asks; what they skip is writing it a second time for readers they will never meet. Capturing the explanation where it was first given works better than asking for a second version.
Permissions that separate who contributes from who reads, so a team can keep its discussions private and still publish the result. Search that works where people already look. Pages that say who decided and who to ask. And a way of getting knowledge in that does not depend on someone finding a free afternoon.
No. SecondBrain 365 writes pages from the conversations a workspace brings it into; the conversations themselves stay where they are. Who can read the pages is set with your existing SharePoint permissions, and only workspace members can send the agent anything.
No. The pages live in SharePoint, in your own Microsoft 365 tenant, and people read them the way they read any other SharePoint page — or ask Microsoft 365 Copilot, which can answer from pages they have access to.