Your team uses AI: how to share the context of your events

Marta is preparing a proposal for the client. She has the notes from the commercial meeting and knows they want a more participatory experience. A few metres away, Diego is working on the production brief in a separate conversation: he knows the venue’s constraints and the time available for the build.

Both use an AI assistant. Each receives a reasonable text based on what they have provided. When they compare documents, the proposal includes an activation that the build cannot accommodate, while the brief retains a format the client already wanted to change.

The agency does not need another person copying notes between chats. It needs work to start from what the project knows, even when that knowledge has come from different people and sources.

Event context is scattered before writing begins

Part of the story lives in the client’s email. Another part is in the call with the venue. Production agreements sit in a document, while the reasons behind a decision remain in a conversation remembered only by those who took part.

When asking an AI agent to do work, you need to provide that information. If each person prepares the context separately, differences appear before the first sentence is written. Marta provides the commercial side; Diego provides the operational side. Neither necessarily knows what is missing.

Even with a good shared folder, someone must locate the current files and explain how they relate. One document might describe an option later rejected. A presentation might contain an earlier budget. Access to files does not ensure a request uses the history that explains the current situation.

Reconstructing that history takes attention the team needs for designing and producing the event. The work returns when a third person asks for a summary or prepares another document.

A shared cloud memory to work from the same project

EVA Brain brings the knowledge received about a project into a shared cloud memory. Team contributions and information from connected sources feed that same base.

Marta can work from her computer while Diego consults the memory from his mobile during a venue visit. The context belongs to the project and remains available for the team’s work. It does not depend on one person forwarding their latest conversation to another.

Brain incorporates the information it receives and maintains its project documentation. The team prepares connections and provides sources; it does not then have to rewrite each email or meeting for that knowledge to become part of the memory.

This changes the request you can make. Rather than reconstructing every constraint yourself before asking for a proposal, you work with an agent that has the accumulated context. You specify the task, the recipient and anything new it should take into account.

Each project’s memory remains separate. A congress’s history should not accidentally become the starting point for another client’s incentive trip.

How team knowledge reaches that memory

Knowledge enters through the sources prepared for the project. You can invite the assistant to team or client meetings so the conversation is transcribed and incorporated into Brain. You can also provide documents and forward emails containing relevant information.

For example, Marta invites the assistant to the meeting where the client explains that the activity should encourage conversations between participants. Diego provides the document containing venue access times and information from the technical visit. Both contributions become part of the event context.

Capture reduces manual documentation, but you need to prepare where each piece of information goes. If a meeting covers two events, identify what belongs to each. If an email describes a possibility, it should remain a possibility until confirmed.

Useful memory includes the facts and the status of the discussion. “The client asked whether we could extend the activity” and “the client approved extending the activity” are different agreements.

The team still makes decisions. Brain collects and organises the knowledge received so those decisions can be used in its work.

Two people, two tasks and the same working base

Return to Marta and Diego. With commercial and operational information incorporated into the project, each can request a different result without explaining the whole story again.

Marta asks for a proposal:

Marta: Prepare a proposal for the client with the participatory activity we discussed. Take the venue’s constraints into account, distinguish confirmed points from pending ones and adapt the document to our brand.

Diego asks for a brief:

Diego: Prepare the production brief for that proposal. Include access times, the build required and the decisions we still need the client to confirm.

Brain uses the same project memory to develop both pieces of work. The proposal can reflect the need for participation and the build constraints. The brief can explain what needs to be produced for that option and identify what remains to be agreed.

The agency gains a common base for review. Rather than comparing two texts arising from different histories, it compares two documents that should reflect the same agreements and constraints.

Generated documents still need review before they are sent or used in production. Check commitments, dates and pending points in particular. More context improves the starting point; it does not remove responsibility for approving the result.

What changes during a working day

The benefit appears in small questions that arise throughout the day. Someone needs to understand why a format was rejected. Another person is preparing a meeting and wants to know what the client expects. Production needs to establish whether a proposal has a constraint not yet reflected in the brief.

With shared memory, those requests can draw on project knowledge without always asking the same person to reconstruct it. You can also request work: develop a proposal, prepare a report or turn available agreements into a brief adapted to your brand.

This helps distribute tasks. Someone preparing a document does not have to have attended every call to access its background. They can consult the context received and concentrate on what they need to produce.

Access from mobiles and computers matters for the same reason. If Diego is at the venue, he should not need to return to the office to check an access agreement. Cloud memory keeps the project available wherever the team works.

The result is not that everyone knows everything by heart. It is that shared knowledge is available when it needs to be used.

Start with one project and recognisable tasks

To prepare Brain, choose an event where context is spread across several people. Identify the sources containing agreements and constraints: client meetings, production documents and relevant emails.

Decide who contributes each source and which task you will test. A proposal and its production brief make a useful pair because they must describe the same decision from different perspectives. You can also compare a client summary with an internal working document.

Review the result with the people who know the project. Has it used both people’s contributions? Does it distinguish a decision from a suggestion? Does it identify missing information rather than filling it in itself?

If you find an incorrect fact or a misinterpreted agreement, correct the memory or the relevant source. Simply fixing the generated text leaves the same problem waiting for the next request.

You can then expand sources and tasks. Each addition should help the team work with context that was previously difficult to bring together.

What to make clear to the team

Brain needs to receive the information. A conversation nobody contributes and a decision existing only in someone’s head do not appear simply because there is shared memory.

It also needs enough context to relate what it receives. If one file replaces another, explain that; if a question remains unresolved, keep it identified. Automatic capture reduces manual documentation, but does not turn ambiguous information into a clear agreement.

Start by preparing that process in a real project. If the team can create a proposal and brief using everyone’s contributions, there is already a tangible change in how it works.

To explore how this could fit your agency or corporate team, tell us how you work. We can start with the sources you already use and the tasks that need the most context.

Frequently asked questions

Do we need to use the same computer or conversation?

No. The memory lives in the cloud and the team works with project context from mobiles or computers. People can request different tasks from that shared base.

Do we have to document every meeting manually?

Brain incorporates information received from meetings, documents and other sources prepared for the project. Initial configuration is required and the team must contribute what does not arrive through those sources, but it does not have to rewrite every conversation manually.

Does Brain only answer questions?

You can also request work using the available context: proposals, briefs and reports adapted to your brand. Review the results before using or sending them.

How do we check whether it helps the team?

Test two tasks that depend on information contributed by different people. Check whether both use the same agreements and constraints, and what context is still missing.