This is an illustrative training example, not a named client engagement.
Situation
A 60-person accounting firm is growing quickly. Partners complain that new clients take too long to become fully operational and staff repeatedly chase missing information. The initial buyer hypothesis is: “We need an AI onboarding assistant.”Diagnostic scope
Business question: where is avoidable effort and delay created from signed engagement letter to client-ready status?Stakeholders
Operations director, onboarding manager, two onboarding administrators, compliance representative, client-facing manager and IT/system owner.Evidence
Iris interviews capture the intended workflow, recurring exceptions and workarounds. Consent-based screen observation with an onboarding administrator reveals repeated copying between systems, manual checks and multiple client-chasing steps.Findings
- Client information arrives through inconsistent channels.
- Staff manually re-key the same details into multiple systems.
- Compliance checks often begin before required information is complete.
- Missing information creates repeated client chasing.
- Senior approval is used for routine exceptions that could be governed by clearer rules.
Opportunity backlog
- controlled client intake;
- deterministic transfer between systems;
- AI-assisted missing-information follow-up with human review;
- clearer exception-approval rules.
Prioritisation
The consultant recommends fixing intake and duplicate entry first. The AI follow-up use case becomes smaller once input quality improves.Business case
The consultant builds the ROI model outside Iris using client-agreed assumptions for annual case volume, manual effort, follow-up frequency and loaded operations cost.Report Composer
The validated process, findings, opportunity sequence and consultant-side business case are assembled into the final report using Iris Report Composer.Executive recommendation
Standardise intake and remove duplicate system entry first. These changes address the dominant evidence-backed causes of delay and create a stronger foundation for later AI-assisted communication.