Role-specific scenarios
Delivered: opportunity areas and use-case designs shaped around executive workflows.
Intended benefit: help leaders recognise where Copilot could be relevant to their own work.
Global bank
Executive AI use-case strategy
From AI possibility to executive relevance
I translated complex executive workflows into practical, role-specific Microsoft 365 Copilot opportunities, supporting a programme for 36 senior leaders at a global bank, including CEOs and other C-suite executives.
My executive-session work covered use-case design, tailored prompts, demonstration scenarios and supporting materials. Separately, I supported the wider programme through learning videos, training and Ask-Me-Anything sessions.
ENGAGEMENT SCOPE
- 36
- Senior leaders supported
- 4 markets
- Singapore · Hong Kong · UK · US
- 144
- Copilot opportunity areas created
The business challenge
A generic Copilot demonstration could show features, but it could not explain which parts of an executive’s work were worth approaching differently.
Leaders across executive management, private banking, legal and communications had different priorities, information sources and recurring decisions. The challenge was to turn that complexity into a focused set of useful, credible AI opportunities for each role.
My contribution
I supported the wider delivery team by connecting executive needs to use cases that could be explained and demonstrated:
- Analysed responsibilities and workflows to identify where information-heavy or repetitive work created friction.
- Assessed opportunities and helped shape four areas per leader, focusing on tasks where Copilot could support review, synthesis or preparation.
- Designed use cases and tailored prompts, defining source requirements, output structures and guardrails to keep the proposed interaction grounded.
- Prepared demonstration scenarios and tailored materials to make each proposed application tangible within the executive’s context.
My contribution was to the analysis, use-case design and preparation supporting the sessions, alongside the team delivering the programme.
The decisions behind the design
01Start with the workflow
I selected opportunities around recurring tasks and decisions, rather than assembling a feature showcase. This gave each scenario a clear business purpose.
02Define the conditions for useful output
I considered the sources, permissions and Copilot capabilities a scenario depended on. Prompts needed to distinguish confirmed information from uncertainty and flag missing or conflicting evidence.
03Make the handoff explicit
I separated information processing from decisions requiring human judgement. This made review responsibilities and the limits of the proposed AI support clear.
Worked example — Executive speaking invitations
Business friction
One executive office expected more than 200 external speaking invitations around a major industry event. The team manually consolidated invitations, identified calendar clashes and weighed competing opportunities before committing.
My design decision
I explored a use case that separated information processing from strategic acceptance decisions. A structured invitation view could support comparison; it could not capture every reputational or relationship consideration.
01Proposed AI support
- Structure event, organiser, audience, topic, date and response deadline.
- Surface overlaps, travel constraints and deadlines using relevant invitation and calendar sources.
- Organise information against criteria such as strategic relevance, audience importance and logistics.
02Human responsibility
- Review the information and resolve gaps or ambiguity.
- Weigh reputation, relationships and strategic priorities.
- Retain final acceptance and commitment decisions.
Value of the design: a clear division of work—AI support for organising and comparing information, with the executive office accountable for the decision.
This was an explored use-case design. The 200+ figure describes expected invitation volume, not invitations processed by Copilot or a measured result.
What this work delivered and enabled
The deliverables made proposed applications concrete. The purposes below describe intended benefits, not measured business outcomes.
Prompts and demonstrations
Delivered: tailored prompts, demonstration scenarios and supporting materials.
Intended benefit: make proposed interactions and outputs easier to understand and discuss.
Practical design boundaries
Delivered: source requirements, feasibility considerations and explicit human-review boundaries within the use cases.
Intended benefit: clarify the conditions and responsibilities needed for appropriate use.
Wider programme enablement
Delivered: learning videos and support for live training and Ask-Me-Anything sessions.
Intended benefit: extend learning beyond the executive sessions and support further exploration.
What I took away
The strongest AI use cases started with understanding someone’s work, not demonstrating a feature.
This project strengthened how I connect business analysis with AI use-case design: understand the friction, shape a practical opportunity and make the proposed way of working clear.