Run the assessment one client account at a time, because each client sets different limits for the same kind of agent. Delivery staff take a test set in the client's industry, and the report shows by concept whether they understand the bounds they will review. It decides the account's first agent and gives the client evidence the people are ready.
Why is an outsourcing provider different?
An outsourcing provider runs operations for other businesses: claims processing for an insurer, collections for a lender, customer service for a retailer. When agents arrive, the provider's staff review agent work under each client's rules, not the provider's own. A refund the retail client allows at depth 3 may be held for a person by the telecoms client next door.
That makes two things harder. Staff must keep several clients' bounds straight, often moving between accounts. And clients want proof that the people handling their customers are trained, usually written into the contract. The assessment addresses both. The head of delivery for the account owns it, with the provider's quality function. For the concept, see what a readiness assessment is.
Who should take it?
The team on one client account, up to 25 people:
- Delivery staff who will approve the agent's held actions.
- Team leaders and quality analysts for the account.
- The account's process owner, who translates the client's rules into working instructions.
Staff who float between accounts should take the test set in the industry of the account they will review most. Who goes first across the provider is discussed in who in an outsourcing provider needs AI agent training first.
The steps
- Person (head of delivery): chooses the client account and agrees with the client which decisions stay with the client's own staff.
- Person (head of delivery): asks for the test set in the client's industry, such as insurance for a claims account or lending for a collections account.
- Person (each member): sits the test alone.
- Person (head of delivery and quality lead): read the report by concept, then share a summary with the client.
- Person (head of delivery, with the client): chooses the first agent and depth for that account.
- Person (process owner): trains the weakest concept using the client's own procedures.
- Agent, depth 1: reads the account's closed cases in shadow, proposing verdicts under the client's limits.
- Person (cohort): reassesses, and the head of delivery shares the result with the client before live work.
| Step | The agent does | Depth | Stays with a person |
|---|---|---|---|
| Choose account | Nothing yet | None | Scope agreed with the client |
| Sit the test | Nothing yet | None | Individual answers |
| Read and share | Nothing yet | None | Summary for the client |
| First agent | Named as candidate | None | Choice, with the client |
| Train the gap | Runs in practice labs | Practice | Learning the client's rules |
| Shadow | Proposes verdicts under client limits | 1 | Comparing with real decisions |
| Reassess | Nothing new | 1 then 2 | Client sign-off for live work |
What does the test usually find in outsourcing teams?
Whose limits apply. In governed terms, every action meets five bounds, and one of them is the customer's own limits. In an outsourcing provider the "customer" is the client. Staff who move between accounts often assume the limits they learned on one account apply on the next. The test, set in the client's industry, shows whether they read which bound decided.
Held versus refused. Delivery staff are used to scripts. When the agent refuses an action because the client's limits exclude it, staff need to know the next step is the client's escalation route, not a workaround.
The record and accountability. The record names who approved each held action. In outsourcing, that person works for the provider while the decision belongs to the client. Staff should understand that their name is on the record and that the client can reconstruct the decision.
What does the report change?
It decides the account, sometimes against the commercial plan. The provider may want to launch on its largest account; the report may show a smaller account's team is ready first.
It shapes rotation. If floating staff score lower on the client-limits concept, the provider can stop rotating them onto that account until they train.
It gives the client evidence. Clients increasingly ask how outsourced staff are prepared for agents. A scored assessment, a decision and a reassessment answer that directly.
What evidence should the provider keep?
The client's never-decide list.
The report by concept for the account.
The first-agent decision, agreed with the client.
The reassessment and the client's sign-off.
What should the head of delivery measure?
- Scores by concept for each account, and for floating staff separately.
- During shadow, cases where staff applied another client's limits.
- After go-live, how many quality findings involve the wrong bound.
Where does it go wrong in outsourcing?
One assessment for the whole centre. A centre-wide average says nothing about any client.
Testing in generic scenarios. Staff working an insurance account need insurance questions; generic ones measure general knowledge.
Losing the evidence when staff leave. Attrition means the assessed team changes fast. Reassess each new intake on the account, and keep the results tied to the account, not to a spreadsheet that leaves with a manager.
Client industries shape the content; see the insurer version and the lender version for two common accounts. The provider's own leadership needs are covered in what AI agent training an outsourcing provider board needs, and the report format in the readiness assessment report template. The topic guide is training teams for AI agents.
Where the Academy fits
Pryme Intelligence Academy offers a free readiness assessment for teams of up to 25, set in the team's industry. Certificates are issued through VerifyDoc.ai and can be checked at source by number or QR code, and a public register lists certified people by name, credential and expiry, so a client can verify an outsourced reviewer's certificate itself.
Questions people ask
Who decides the first agent, the provider or the client?
Both. The client sets the limits and owns the decisions; the provider knows its team's readiness. Agree the choice jointly and record it.
Should the client's own staff take the test too?
If they will review escalations from the provider's team, yes, in their own sitting. Both sides then share the same vocabulary.
How often should an account reassess?
At each new intake, before any rise in depth, and whenever the client changes its limits.