Leadership
Founders, C-suite, department heads
You set the direction and carry the risk, and you are being asked to approve tools and budget without a clear picture of what any of it returns.
Usually starts with
Services · AI training
AI training for teams that actually use it
Good AI training teaches your people to use AI on their own work, not just what it is. Traq Collective runs six hands-on tracks covering AI strategy and governance, Claude, ChatGPT, Microsoft 365 Copilot, Gemini and AI agents, each rebuilt around your team’s real tasks. In person across the UAE, remote worldwide.
Last updated:
workshop plan
- 1. Learn the real work
- 2. Hands-on with your tools
- 3. Habits that actually stick
Who we train
Three rooms, three different problems
The board question and the shop-floor question are not the same question. We split the catalogue by who is in the room, then pick the track from there.
Champions
Managers and the people others already ask
You will be the one the team comes to after we leave, so you need more depth than everyone else and a way to teach it on.
Usually starts with
Everyone else
The people doing the daily work
You have the licence, you have opened it twice, and nobody has shown you how it fits the job you actually do.
Usually starts with
The catalogue
Six tracks. Pick the ones your team needs.
Every track is rebuilt around your work before we deliver it, so the modules below are the spine rather than the script. Open any track to see what is in it.
Tools we put hands on:ClaudeChatGPTMicrosoft 365 CopilotGeminiNotebookLMPerplexity
AI Strategy & Governance
Leadership- For
- Founders, C-suite and department heads
- Format
- 90-minute exec taster, or half day
The session to run before you spend anything else. Leadership leaves aligned on what AI realistically does for this business, which work to point it at first, and the rules everyone else will be held to.
They leave with
A ranked use-case shortlist, a draft acceptable-use policy, and a 12-month roadmap with owners, budget and the numbers you will report on.
See the 6 modulesHide the modules
What AI can and cannot do for your business
Separating real capability from the marketing, using your industry and your workflows as the examples.
Finding the use cases that pay
Scoring candidate workflows on hours saved, risk and effort to build, then ranking them so the first three are obvious.
Data, privacy and where your information goes
Which tools train on what you type, what belongs in them, what never does, and how the UAE picture differs from the US and EU.
Writing an acceptable use policy people follow
Approved tools, review steps, what always needs a human sign-off, and how to say it in one page rather than twelve.
Licensing and what you already pay for
Seats, tiers and the overlap between ChatGPT, Copilot and Gemini that most companies are quietly buying twice.
The 12-month roadmap
Sequencing, owners, budget and the measures that tell you it is working, set out quarter by quarter.
Applies across all tools
Claude for Business
Core- For
- Anyone whose week is documents, analysis or writing
- Format
- Half day, or full day using your own files
Claude is the strongest tool most teams have never properly opened. This track puts it on your contracts, tenders, reports and research, and sets up the parts that make it repeatable rather than a one-off chat.
They leave with
A shared Project holding your business context, a starter prompt library, and two or three recurring tasks rebuilt so they run in minutes.
See the 6 modulesHide the modules
Briefing Claude like a capable new hire
The structure that gets a usable answer first time: context, task, constraints and an example of good.
Projects
Loading your standing context once so every conversation starts already knowing your business, your clients and your tone.
Long documents at full length
Contracts, tenders, board packs and research read end to end, with the passages quoted back so you can check them.
Artifacts
Drafting documents, tables and small internal tools you can edit in place, keep and hand to someone else.
Skills
Teaching Claude a process once so the whole team runs it the same way every time, instead of everyone prompting differently.
Checking the work
Where Claude is strong, where it is weak, and how to spot an answer that is confidently wrong before it reaches a client.
ClaudeClaude ProjectsArtifactsSkills
ChatGPT for Business
Core- For
- Whole teams, especially sales, marketing and admin
- Format
- Half day, or full day using your own work
The foundation track. Most people use ChatGPT as a better search box and stop there. This one covers the features that turn it into something the whole team can run a repeatable job on.
They leave with
Custom instructions set for each person, at least one Project per team, and a custom GPT built live for a job you do every week.
See the 6 modulesHide the modules
Prompting that holds up under pressure
Structure, context and worked examples, and why the answer quality is usually your brief rather than the model.
Projects and custom instructions
Setting things up so you stop re-explaining your company at the start of every conversation.
Custom GPTs
Packaging a repeatable job so a colleague can run it well without being good at prompting.
Files, spreadsheets and analysis
Pulling numbers out of PDFs and workbooks, and running the analysis you would otherwise queue for someone else.
Research you can trace
Getting sourced answers, checking them, and knowing which questions the tool should not be trusted with at all.
The parts of the toolkit nobody opens
Voice, images, screen and file handling, and where each genuinely saves time rather than being a demo.
ChatGPTProjectsCustom GPTsData analysis
Microsoft 365 Copilot
Core- For
- Teams already licensed for Microsoft 365
- Format
- Half day
You are probably already paying for this per seat. Copilot behaves differently in every app it sits in, which is why adoption stalls, so this track goes app by app and is blunt about where it is worth your time.
They leave with
Every person leaves with Copilot working on their real inbox, their real documents and their real spreadsheets, plus a clear rule for which job goes to which tool.
See the 6 modulesHide the modules
Where Copilot actually lives
A tour of Copilot across Word, Excel, Outlook, Teams and PowerPoint, and an honest read on which ones earn their keep.
Outlook and Teams
Triaging a full inbox, catching up on a meeting you missed, and drafting replies that sound like you wrote them.
Word
First drafts built from documents you already have, rather than from a blank page and a vague prompt.
Excel
Formulas, cleanup and analysis described in plain language, including the cases where it still gets it wrong.
Grounding on your own files
How Copilot reads SharePoint and OneDrive, why permissions decide what it can see, and what that means for confidential work.
Copilot, ChatGPT or Claude
A simple rule for routing each job to the right tool, so you stop paying for three and only ever using one.
Microsoft 365 CopilotOutlookTeamsWordExcelPowerPoint
AI Agents & Automation
Advanced- For
- Ops leads and whoever owns a process end to end
- Format
- Full day, or a 4-week build programme
For teams past the basics who want work happening without a person driving it. We build a real agent on one of your own processes during the session, rather than demoing someone else’s.
They leave with
One live agent running a real process, a map of the next three worth building, and the approval gates that keep a human in the loop.
See the 6 modulesHide the modules
Automation, AI and agents
What the three words actually mean, and which one your problem needs, because it is usually the cheapest of the three.
Map the process before you automate it
Drawing the current process honestly, because automating a broken one just produces broken output faster.
Building your first agent
Triggers, steps, tools and the handover back to a person, built live on a process your team runs every week.
Human in the loop
Where to put approval gates so nothing reaches a client unchecked, and how to keep them from becoming a bottleneck.
Connecting your stack
Wiring the agent into email, CRM, sheets and storage, using the accounts and permissions you already have.
Running it on a Tuesday
Monitoring, failure, cost and ownership, so the thing still works in month six when nobody remembers building it.
ClaudeChatGPTZapierMakeYour existing stack
Also delivered on request
- Gemini & Google Workspace — Teams running on Google Workspace. Half day.
Need something that is not here? Most engagements end up as a track built from scratch around one team’s workflow. Tell us what the work is and we will scope it.
How it runs
What a Traq workshop actually looks like
The session in the middle is the part people picture. The two either side are the part that decides whether any of it is still happening in three months.
Before: we learn the work
A short survey of the people attending, a call with you, and a look at how a few real tasks get done today. The agenda gets built from that. Nothing is pulled off a shelf.
During: hands on your own tasks
Nobody watches a demo. People bring live work and leave having done it faster, with the guardrails explained as we go rather than as a policy slide at the end.
After: the part most training skips
A recap pack and prompt library go out the same week, and we come back a few weeks later. Research puts the tipping point at five-plus hours of practice, so one session on its own was never going to hold.
Deliverables
What your team walks away with
Not a certificate. The working setups, the decision artefacts, and the follow-up that stops it fading.
In their hands the same week
Everyone leaves with working setups rather than notes to action later.
- A prompt library written for their role, not a generic list
- A Project, custom GPT or Skill set up on your business context
- Two or three of their own recurring tasks rebuilt and timed
For the business
The artefacts leadership needs to make the next decision without another discovery round.
- A ranked use-case shortlist with hours attached to each one
- A draft acceptable-use policy covering tools, data and sign-off
- A 30, 60, 90 day plan with owners and the measures you report on
So it holds after we leave
One session changes a mood. Keeping the habit takes a bit more than that.
- A recap pack and the session recording, sent the same week
- A named internal champion briefed to answer the follow-up questions
- A return session a few weeks later, once the friction has surfaced
Formats
How long it takes and how many can attend
Any track runs in any of these formats. Most teams start with one workshop, then roll the rest out over a few weeks once they have seen it land.
| Format | Length | Group size | Best for |
|---|---|---|---|
| Exec taster | 90 minutes | Up to 30 | Getting leadership aligned before you commit budget to anything |
| Workshop | Half day | 4 to 25 | One team, one track, one tool, taken properly |
| Deep dive | Full day | 4 to 20 | Hands-on rebuilding of the work your team does every week |
| Programme | 4 to 6 weeks | Any size | A whole-company rollout, track by track, with follow-ups built in |
- In person
- Dubai, Abu Dhabi and across the UAE, at your office or a venue we arrange.
- Remote
- Worldwide, live rather than recorded, with the same hands-on format.
- Lead time
- Typically two to three weeks, because the prep before the session is real work.
Compare
AI training vs leaving your team to figure it out
What changes when training is guided and built around your work, versus hoping people pick it up on their own.
what doesn’t work
Figuring it out alone
- A few keen people try it; most never start
- Generic tips from YouTube that rarely fit the job
- Ad-hoc use with no rules, so people lose trust fast
- Months of trial and error, often abandoned
- Early enthusiasm fades once the novelty wears off
what works
Guided AI training
- Most of the team uses AI on real work within weeks
- Built around your roles, tools and actual tasks
- Clear guardrails for accuracy, privacy and review
- Hours saved start showing in the first month
- Habits reinforced with follow-up so they hold
FAQ
AI training: common questions
Short answers on the tracks, the tools, the group sizes and how delivery works.
Which AI tools do you train on?
Claude, ChatGPT, Microsoft 365 Copilot and Gemini, plus the automation tools that sit around them. We train on what you already pay for: if you hold Microsoft licences we go deep on Copilot, and if you run on Google Workspace we take the Gemini track. Most teams leave with a simple rule for which tool gets which job.
How long does AI training take?
Most teams start with a half-day or full-day workshop, then a short follow-up a few weeks later so the habits stick. Research points to five-plus hours of practice as the tipping point for regular use, so we scope enough hands-on time to get there, not a one-off talk.
What is in an AI strategy and governance session?
It is the leadership track. We work through what AI realistically does for your business, score and rank the use cases worth funding, agree what data can go into which tool, draft an acceptable-use policy in plain language, and finish with a 12-month roadmap carrying owners, budget and the measures you will report on.
Do you train non-technical teams?
Yes. Most of the people we train are not technical. We teach AI on the work they already do, in plain language, with no code required. Finance, sales, operations and admin teams are exactly who benefit most from practical, role-specific training.
Can you build a track around our own tools and processes?
Yes, and most engagements end up there. The six tracks are a spine, not a script. Before delivery we survey the people attending and look at how a few real tasks get done today, then rebuild the agenda around that. If your work lives in a tool we have not listed, we still train on it.
What size group works best?
Hands-on tracks run best between four and twenty-five people, because everyone needs to be doing the work rather than watching someone else do it. The 90-minute leadership taster scales to a couple of hundred. For a whole-company rollout we run the same track several times instead of putting everyone in one room.
Do you run AI training in the UAE?
Yes. We run in-person workshops for teams across Dubai, Abu Dhabi and the wider UAE, and deliver the same training remotely for teams worldwide. The work is the same wherever your team sits; only the delivery format changes.
Book a call
Want your team using AI by next month?
Book a free call. We will scope a track around your team’s real work. No deck, no obligation.
Other services
The rest of what we do
Most teams combine two or three of these into one engagement. Here are the others.