The best AI projects start with a real problem and a team that understands it. A well-run use-case workshop turns scattered ideas into a ranked shortlist with owners, risks and next steps. This guide sets out a format you can run yourself or ask us to facilitate.
Who to invite
Bring the people who do the work, not only managers: frontline staff, process owners, someone from IT or data, and a representative from risk, privacy or security. An executive sponsor who can say yes to the outcome keeps the workshop honest.
Start with problems, not technology
Ask participants to list tasks that are repetitive, slow, error-prone, rule-based or document-heavy. Capture who does them, how often and what a good result looks like. Resist starting from a tool name. Technology comes after the problem is clear.
Shape each idea consistently
Use a one-page template: the problem, the user, the data needed, the type of AI that might help, what a person still decides and what success looks like. Consistency makes ideas comparable.
Score for value, feasibility and risk
| Lens | Questions |
|---|---|
| Value | How much time, cost or risk does it change? How many people benefit? |
| Feasibility | Is the data available and good enough? Does an existing tool already do it? Who will own it? |
| Risk tier | Does it affect people’s entitlements? Does it use sensitive data? Does it act autonomously? What happens if it is wrong? |
Plot ideas on value against feasibility, then mark risk tier. Early wins are usually high value, feasible and low risk, such as summarisation, drafting and search over approved documents. Higher-risk ideas are not rejected, but they get a governance path and a pilot designed with human review.
Leave with decisions
The output should be a ranked shortlist of three to five use cases, each with an owner, a risk tier, the data and permissions needed and an agreed next step such as a pilot, a data fix or a policy decision. Add them to your AI register straight away.
Follow through
Review the shortlist with leadership, pilot one or two use cases with clear measures and keep the rest in a backlog. See our guides on AI governance in practice and AI for South Australian councils.
Book a session
Our team facilitates AI use-case discovery workshops on-site in Adelaide or online, tailored to your team’s work and tools. Book a session or explore Training & Team Consultations. For governance of the use cases you choose, see AI Governance & Assurance.
This is general guidance, not legal advice.