Work with clients

A decision from a client meeting, a technical test, a question still waiting for an answer: when these are scattered across Slack and meeting notes, every handoff starts with someone explaining the history again.

FDEs (Forward Deployed Engineers), implementation specialists, consultants, and support teams can use an Epismo case to keep the client’s goal and the work’s history together. The next person or connected AI agent reads that case to continue the work. Methods worth repeating can become playbooks for future work.

This guide follows an illustrative engagement with Client A: pilot AI with ten salespeople, handle a request for CRM integration, and check results before recommending rollout. At the end, leave the client a method they can use to keep preparing sales meetings themselves.

Keep the history in a case and the method in a playbook

A case is a place to move one piece of work forward. In “Evaluate Client A’s sales AI pilot,” collect the goal, agreed scope, meeting outcomes, tests, and decisions. When someone takes over, they can see what has been decided and what remains open.

A playbook holds steps you will use again. Your internal “Implement and evaluate a sales AI pilot” method can guide the next client engagement. The client’s sales-preparation playbook serves a different audience: salespeople doing their daily work. Give each playbook a clear purpose and audience.

Start with one engagement case and the internal team using it. Add client collaboration and playbooks as the work calls for them.

1. Keep goals and agreed scope in a case

When a client says “We want to use AI,” ask what would count as an improvement. For Client A, start with the time spent preparing meetings and the quality of proposals.

Create “Evaluate Client A’s sales AI pilot” and record the context below. Include unanswered questions so they become things to confirm in the next meeting.

If you have an “Implement and evaluate a sales AI pilot” playbook, start the case from that version. Otherwise, begin with a title and develop the method as you work.

For creation steps, see First case.

2. Give the delivery team access

The implementation specialist talks to the client, a technical specialist tests the connection, and the delivery lead decides on the next proposal. To give them the same context, share the engagement case with your internal team as editors and assign someone to lead it.

You can start internally. Recording what you learn in client meetings makes that context useful to your team before the client starts using Epismo.

When the client will work with you in Epismo

People from different workspaces can work in one team. For example, “Client A delivery” lets both companies participate from their own workspaces and manage their own participants.

1. Ask your workspace admin to create the shared team and arrange a team invitation to the client contact.
2. The contact accepts and selects their workspace, then the client adds the participants they need.
3. Share the cases and playbooks you will use together with the team. Give it editing access for joint updates or viewing access for reading.

You can also invite client members into your workspace and add them to the team. Use this option when your company will manage everyone’s membership.

Keep internal commercial discussions in internal cases, and choose what to share with the client. Before starting, check together that a client participant can open the shared case.

For invitation steps, see Workspaces and teams. For permission details, see Access and identity.

3. Leave records the next person can act on

During the pilot, the client asks, “Can we also connect it to our CRM?” Saving that sentence alone leaves the next person unsure whether to start work.

Record the context, current decision, open questions, and next action. Link supporting Slack threads or Google Drive files as external sources.

What changed: The client requested CRM integration.
Evidence: The client’s Slack request and the technical specialist’s connection test.
Decision: Continue the ten-person pilot; start integration after agreeing scope and timing.
Open questions: Data to integrate, extra time needed, and client approver.
Next action: The implementation lead confirms these in the next meeting and records the agreement.

Start with the request and open questions, then add results as tests and decisions progress. The technical specialist reads the case to continue testing; the delivery lead uses those results to decide on the next proposal. Someone taking over can follow the same history.

With AI connected, you can ask: “Read Client A’s case and recent records. Summarize the agreed scope, open questions, and next actions for our next meeting.”

For record and handoff operations, see Coordinate a case.

4. Turn scope changes into decisions

Testing shows that CRM integration needs additional work. Record the work required, timing impact, and options in the case. Assign any remaining investigation as tasks.

For example, record a proposal to “Add CRM integration and extend the pilot by one week.” Create an approval task tied to that record and assign it to the reviewer. Save the agreed scope and dates after the decision so a later reader can see which proposal was approved and what to do next.

For tasks and approvals, see Coordinate a case.

5. Check results against the client’s goal

Once the pilot works, return to the original goal. Client A needs to know how much easier meeting preparation has become and whether proposals have improved.

Record the pilot results and request AI review to examine the case evidence against its goal. Read the report to assess gaps and decide what to check next.

For Client A, record preparation-time and proposal-quality measurements before and after the pilot. The delivery lead considers the review and measurements before recommending rollout. Request review when needed; a person decides which proposals to adopt.

6. Carry lessons into the next engagement

If a lesson stays only in Client A’s case, the next engagement may repeat the same missed check or manual work. Decide what would make future work easier and send the finding to the place where it can be addressed.

Submit method improvements as playbook suggestions. Editors consider them and publish an updated version. Remove client-specific details when contributing to a method used across clients.

For operations, see Improve a playbook.

7. Leave the client a method they can keep using

Leave the client something they can use in daily sales work after the engagement ends. Their sales-preparation playbook should explain which sources to use, what AI does, and where a person checks the work. For example:

1. A salesperson provides the meeting goal and checks CRM history and prospect information.
2. AI uses those sources to organize proposal hypotheses and questions.
3. The salesperson verifies sources and proposal criteria, then finalizes the proposal.
4. The team records the outcome in a case and considers improvements.

If your company has a reusable “Research a prospect and prepare a proposal” playbook, copy it and adapt it to Client A’s CRM, products, and proposal criteria. The source playbook and version remain traceable, while client changes are managed independently. Otherwise, create a new playbook.

Check the content is suitable for the client, publish a version, and share it with the team that will use it. If the client will own and maintain it, create or copy it in the client’s workspace.

After sharing the playbook, have the client prepare one real meeting with their own tools and connected AI. Check that they can record results, know where human review is needed, and have someone responsible for updates and a support contact. Record the handoff results in the engagement case.

For copying, see Find and use a playbook. For creation, publishing, and sharing, see Create and publish a playbook.