Better instructions
Brief AI like a colleague
A vague request gets a confident answer built on guessed scope and unstated assumptions.
You'll build: A reusable task-brief template: outcome, context, limits and how the answer will be judged.
Enterprise AI Academy
Your work.Your context.Your AI.
A practical, checkable way for your team to work with AI, taught in eight 30-minute workshops.
What is the Enterprise AI Academy?
The Enterprise AI Academy is a programme of hands-on workshops for business teams, designed and developed by Digitalis. It teaches people to work with AI the way they would with a capable colleague: brief it properly, give it the right context, check what comes back and keep people in charge of anything that matters. Each workshop teaches one method, applies it to a realistic task and finishes with something participants can keep using. The same methods are taught on each supported platform, and they're designed to carry over when your tools change. Exercises use fictional case material by default, and where a workshop invites participants to use their own writing, we ask them to use only non-confidential samples their organisation allows.
Eight workshops. One progression.
What each session is built to produce. Results depend on the platform features your organisation has turned on.
Portable context
Own your AI brain
Decisions and working preferences end up scattered across chat histories, so changing tools means starting again.
You'll build: A portable context pack and a structured handoff designed for another AI to pick up.
Personal voice
Write like yourself
Generic AI polish flattens your voice, and every new chat needs your preferences explained again.
You'll build: A personal writing profile, plus a list of patterns you want kept out.
Source-grounded work
Work with evidence
AI summaries blend facts, guesses and out-of-date material into one confident paragraph.
You'll build: A status brief that links each claim to a dated source.
Decision quality
From information to decision
AI comparisons weigh every option evenly and stop short of saying which one holds up.
You'll build: A decision brief with clear criteria, a conditional recommendation and the trade-off behind it.
Repeatable work
Build a reusable assistant
A good weekly report gets rebuilt from a blank chat every week, with uneven results.
You'll build: A reusable weekly reporting assistant, with notes a colleague can follow.
Supervised execution
Delegate with control
"Take care of this" blurs the line between preparing something and acting on it.
You'll build: A supervised workflow that maps where AI prepares the work and where a person approves it.
Evaluation
Challenge and verify AI
A persuasive answer can hide a stale source, a wrong number or a missing approval.
You'll build: An evaluation checklist and five tests designed to be rerun when models or tools change.
Who is it for, and how does it run?
- Who it's for
- Business users, managers, operational teams, analysts and executives, in areas such as operations, policy, customer experience, programme delivery, communications, finance, people and culture, and executive support.
- Session length
- 30 minutes per workshop, with an optional 15 to 20-minute discussion after each.
- Platforms
- Microsoft 365 Copilot, Claude or Gemini, depending on what your organisation has enabled. Each team works on one platform.
- Adapts to your licence
- If your organisation has switched off some features, each session has a plain-chat alternative.
- Technical skills
- Everything happens in plain language, in chat and simple text files.
- Each session
- A guided exercise and a quick quality check, designed to finish with a saved piece of work.
- Delivery mode
- Online or in person
Not affiliated with or endorsed by Microsoft, Anthropic or Google.
How the workshops work
Synthetic data first
Teams practise on fictional case material, and we ask participants not to put confidential business information into an AI tool during training.
Supervised action
We teach a pattern where AI prepares the work and people approve anything that is sent, published or changed.
Verification
Teams learn to check outputs against clear tests before relying on them.
A look inside
Title slides from two of the eight workshops.
Talk to us about a workshop for your team
Tell us a little about your team, which AI platform you use and what you'd like people to do better with it. We'll get back to you to talk it through.
Digitalis is an independent Australian business. Our workshops are not affiliated with, sponsored by or endorsed by Microsoft, Anthropic or Google. Microsoft, Microsoft 365 and Microsoft 365 Copilot are trade marks of the Microsoft group of companies. Claude is a trade mark of Anthropic, PBC. Gemini is a trade mark of Google LLC.