Claude Training for Teams
Hands-on Claude training built around the work your team actually does. Three half-day levels, from AI fluency to automation, so the licences you are paying for start earning their keep.
Who this is for
Organisations running Claude on a paid plan (Pro, Team, or Enterprise) who want their people to use it properly rather than occasionally. Typical buyers are heads of People and Culture, L and D leads, IT and digital leaders, and operations managers.
If you have rolled out Claude and adoption has been patchy, that is a training gap rather than a tooling gap. The seats are bought and the spend is committed. What is missing is the point where someone opens Claude on a Monday morning because they already know what to do with it.
We also work with teams evaluating Claude who want their people capable before the licence spend scales.
What your team will be able to do
- ✓
Explain in plain language how Claude produces an answer, and why that explains both its strengths and its failure modes
- ✓
Write prompts that give Claude enough context to be useful, then iterate when the first attempt falls short
- ✓
Review Claude's output critically, including catching the answer that reads well and is wrong
- ✓
Apply your organisation's rules about what can and cannot go into an AI tool, and understand the reasoning behind them
- ✓
Set up Projects that carry instructions and knowledge, and share them so the team works from the same base
- ✓
Produce finished documents, analysis and decks rather than answers they then rebuild elsewhere
- ✓
Recognise which recurring work should be handed over to Claude, and which should not go near a model at all
The training levels
Each level is a standalone half-day. Most teams run them in sequence, with time in between to put the last one to work. The content is set in discovery around the work your cohort actually does.
Level 1: AI Fluency with Claude
A shared foundation for a mixed room: how AI works, why it gets things wrong, what is safe to put into it, and the judgment to review what comes out. The thinking is tool-agnostic and holds whatever tool your people open next. The practice is in Claude, on your real work.
Modules
Level 2: Practical AI with Claude
Out of the chat box and into the way work actually gets done. Projects that hold your context so nobody rebuilds it every time, Artifacts that turn answers into finished work, and connectors that bring your files and tools in. We go deep on the surfaces your people will actually use.
Modules
Level 3: Automations and Agents with Claude
Level 2 makes one person faster at one task. Level 3 is where work starts running without them. Three ways to hand work over, and the judgment to choose between them. Most ideas that arrive as "let's build an agent" turn out to be something simpler, and sorting that out early is where the value is.
Modules
We would not recommend starting at Level 3, which is what most organisations are currently doing and the reason their AI programs stall. Each session scales whatever judgment your people already have, so building on a gap automates the gap.
Why fluency comes first
You cannot build on judgment your people do not have.
Skip fluency and you scale mistakes, not capability. A team that does not know why AI gets things wrong will not check it. A team that does not understand what Claude can see will put data where it should not go. Build automations on top of that and you have automated judgment your people never had.
Licences you are already paying for.
The gap between paying for a licence and someone using it every week is a training gap, not a tooling gap.
Pilots that do not convert.
A pilot proves the technology works. It does not prove your people will use it. Adoption is where most AI programs quietly stall.
Output nobody trusts.
When people cannot tell when AI is wrong, they either accept everything or reject everything. Both are expensive.
Data in the wrong place.
People who do not understand what a tool can see cannot make good decisions about what to put into it. That is a governance problem you inherit later.
Format and pricing
One cohort, one level
From $5,000 + GST
One cohort, any two levels
From $9,500 + GST
One cohort, the full pathway
From $13,500 + GST
No headcount limits.
One session fee covers the room, whether that is eight people or thirty. What changes with group size is the hands-on format: larger groups shift toward demonstration and guided practice rather than individual exercises. Either way, the content is tailored to your team's actual use cases.
What counts as a cohort.
One team, function, or department whose people work on the same kind of problem. That shared context is what lets us build the session around their real use cases. For organisations with diverse teams, we recommend separate sessions by business unit rather than one large all-hands, because different teams have very different use cases.
Why the pathway costs less than three separate sessions.
It is not a volume discount. In the first engagement we are still learning how you work. By the third we know your workflows, your tools and your people, so more of each session is spent building on what came before.
Delivery: In-person or live online, Australia wide and globally.
Prices are AUD, excluding GST. Travel billed at cost.
What you own afterwards
Everything we build is yours to keep and reuse.
How we build your session
01
Discover.
A short discovery call with your leads to map real workflows and gauge the range of starting levels across the team.
02
Design.
A custom session built around those tasks and your tools, branded for your organisation.
03
Deliver.
Live onsite or virtual, hands-on, with tasks pulled from your own work rather than ours.
04
Embed.
A prompt library, quick-reference pack, and 30-day follow-up so the fluency lasts.
Discovery is a build step, not a sales call. It is why the prompt library you keep is made of your workflows rather than our examples.
Frequently asked questions
What is Claude training?
Claude training teaches teams to use Anthropic's Claude for their day-to-day work. Our program runs as three half-day levels: AI fluency, practical use across Claude's Projects and Artifacts, and automation with Cowork and skills. Each session is built around your team's real workflows rather than generic examples.
How long does Claude training take?
Each level is a half-day, around four hours. Teams can take one level or run all three as a pathway, usually with a few weeks between sessions to put each one to work.
How much does Claude training cost in Australia?
Pricing starts at $5,000 + GST for one cohort at one level, $9,500 for two levels, and $13,500 for the full three-level pathway. We price the cohort, not the person, so bringing eight people or thirty does not change the fee.
Do we need Claude licences before the training?
Yes. Participants need an active paid Claude plan (Pro, Team, or Enterprise) to do the hands-on exercises. We confirm what your organisation has enabled during discovery and shape the session around it.
Which level should we start at?
Most organisations should start at Level 1, including many who think they should not. Daily use is not the same as competent use. If your people can already explain why AI hallucinates, what Claude can and cannot see, and where the guardrails sit, we start at Level 2.
Do you deliver Claude training onsite in Australia?
Yes. We deliver onsite in Sydney, Melbourne, Brisbane and Perth, and live online for distributed teams. We also deliver across New Zealand, Singapore and Malaysia.
How is this different from Anthropic's own free course?
Anthropic's course teaches an individual, at their own pace, using generic examples. This is a live session for a whole team, built around your workflows, your data policies and your approved tools, so the room leaves at a shared baseline rather than a certificate.
Can Claude work inside Microsoft 365 or Google Workspace?
Yes. Claude has add-ins for Word, Excel, PowerPoint and Outlook, and connectors for Google Drive, Gmail and Calendar. Which surfaces we cover in the session depends on what your organisation runs, confirmed in discovery.
We already use Copilot or Gemini. Does Claude replace them?
Usually not. Most organisations end up running more than one AI tool, and the useful question is which tool suits which work. We cover that in discovery rather than assuming you are switching.
What if our team has very different roles?
We run separate cohorts by team or function rather than one large all-hands session. Different teams have genuinely different use cases, and building the session around a specific group's work is what makes the training transfer.
Is this training suitable for non-technical staff?
Yes. No technical background is assumed and there is no coding in any of the three levels. The training is for people who write, analyse, plan, advise and decide.
Can you train our whole organisation?
Yes. Where teams have genuinely different workflows, each becomes its own cohort, because that is what lets us build around their actual work. One broad session for everyone is cheaper. Several tailored ones are what actually transfer.
Last updated: August 2026
Related workshops
Ready to get your team using Claude properly?
30-minute discovery call. We work out where your people actually are, which is rarely where the org chart says they are, and which level to start at. You will get a straight answer, including if that answer is that you do not need us yet.