Show clients what their automation actually did
Make the agents your team operates easier to investigate, review, and support with a shared operational record.
Your client asks what happened after the workflow went live.
An automation handoff needs more than a happy-path demo. Make the ongoing work reviewable and agree which decisions remain with the people responsible for the result.
Scope the responsibility
Document the workflow, the runtime owner, the reviewer, and which external actions require a person’s decision.
Make support inspectable
Report run histories and outcomes. Investigate incidents from the available context and keep reviewer decisions distinct from agent claims.
Evaluate before expanding
Pilot one client workflow. Review the operational record together and agree support scope and commercial terms before a broader rollout.
The conversation after go-live
The demo went well and the automation is live. Then the client asks what it actually did last month, why the Tuesday report was late, and who decided it could email their customers. An agency that can answer those questions with a record, rather than a reassurance, keeps the account.
One record per client
Keep each client’s agents in their own workspace so their runs, incidents and reviews stay separate.
- Run history and outcomes for every automation you operate.
- Incidents for missed schedules and stalled runs, routed to your team first.
- Reviews showing a person checked the deliverable before it went out.
- Approvals showing who allowed each consequential action, and when.
Agree responsibilities before launch
Write down, per workflow, who owns the runtime, who reviews the output and which actions need a person’s approval. Configure exactly that: reviewers get the outcomes, approvers get the approval requests, and the client can see that the rules are followed.
Price and scope support with evidence
Run history turns support from guesswork into a scope. You can see how often each automation needs attention, how long incidents take to resolve and which clients generate the most review work. That is the evidence for a retainer, or for a conversation about a workflow that needs rebuilding.
How to measure the pilot
Track which workflows have a named reviewer, how many delivered outcomes were accepted, and how long incidents took to investigate. These are evaluation measures, not promised results.
Explore the relevant capability →Frequently asked questions
Can clients see their own agents?
You can invite client stakeholders to their workspace as members, so they see the same runs, incidents and reviews your team does. Owners and admins keep control of settings and approvals.
How should an agency structure workspaces?
One workspace per client keeps records, alerts and member access separate, and makes it simple to hand a workspace over if the engagement ends.
Is there pricing for agencies with many clients?
Each workspace has its own plan, from Free for small pilots to Business for 100 agents. For many client workspaces or specific terms, talk to us about Enterprise.
Related guides
A step-by-step checklist for deploying an AI agent to production: define a good result, test failure paths, set limits and approvals, then decide go or no-go.
Read the guide Approving an AI agent’s action does not prove it happenedA person approving an AI agent’s action does not show it happened correctly. How to record authorization, receipts and independent verification separately.
Read the guide What an AI agent audit trail should recordWhat an AI agent audit trail should record: runs and steps, claimed versus measured data, who approved each action, policy and credential changes, retention.
Read the guideConnect your first agent
Create a free workspace, connect the agent you already run, and watch its next run arrive. Free covers 3 agents, with no card required.