Claude and Codex for your team’s real work
Hands-on workshops for leaders, business teams and engineering teams. You leave with templates, usage rules and workflows for real tasks, with clear human oversight.
The problem we solve
AI tools often begin as individual experiments: one person writes emails, another analyses documents and someone else tests code. The business has no shared rules, safe examples or processes with a visible benefit. Training needs to start with real work, rather than a list of tricks.
Who this is for
- Owners and leadership teams who want to introduce AI without confusion around tools and data
- Sales, marketing, administration and operations teams with repetitive writing, research and reporting
- Engineering teams who want to use Claude Code and Codex for safer repository work, review preparation and documentation
- Businesses that need clear rules before bringing AI tools into daily work
What you receive
Sample workshop agenda
Using one safe task, we work through preparing instructions, drafting and human review. This is an illustrative agenda. We agree the actual scope, duration, format and required access to suit your team.
Choose a safe task
We record what you do, the result you need and how you will check it. We use fictional or approved data for the exercise.
Prepare instructions and context
We specify the sources, the required output format and what the tool must mark as unknown. We identify which actions remain with a person.
Check the output
We compare the draft with its source. We check facts, names, deadlines and omissions, and refer unclear points back for confirmation.
Save a repeatable process
We save the instructions, checklist and review rules so the team can repeat the task and recognise when help is needed.
What to prepare
Choose one recurring task and a safe example. Remove personal, confidential and unnecessary data.
Check which accounts and tools you are allowed to use. We agree the required access, any subscriptions and the format of the materials before the hands-on work.
Examples by role
Business team can practise a customer reply or meeting summary, checking the facts and tone.
Engineering team can prepare a code-change review or documentation, checking the sources and tests.
An example of what you take away
A checked meeting summary with decisions, next steps and sources, together with instructions and a checklist for repeating the task. In this example, the owner and deadline for the new-form instructions remain unagreed because the notes do not specify them.
How the workflow works in practice
Illustrative examples with fictional data. They show the process and human review, not the results of a client project.
Turn meeting notes into decisions and next steps
- Input
Fictional notes, 28 October 2026. [1] Ana: I will collect customer feedback by 30 October 2026. [2] Marko: We will keep the current price until the next review. [3] A draft of the instructions for the new form needs to be prepared. No owner or deadline has been agreed.
- Output
DECISION Keep the current price until the next review. Source: [2]. NEXT STEPS 1. Collect customer feedback. Owner: Ana. Deadline: 30 October 2026. Source: [1]. 2. Prepare draft instructions for the new form. Owner: not agreed. Deadline: not agreed. Source: [3]. TO CONFIRM Who will prepare the form instructions, and by when?
- Process
- Separate confirmed decisions from tasks and open questions.
- Include an owner, deadline and source for each item only when they appear in the notes.
- Request human confirmation before transferring anything to a task tool.
- Human review
- Compare every decision and task with its cited source.
- Confirm names, deadlines and whether a decision was actually made.
- Confirm unassigned responsibilities and deadlines with the team. AI does not assign them.
- What we measure
- Time from notes to a reviewed summary.
- Number of corrections, missed items and open questions before sharing.
Why DTC
Training uses safe or anonymised examples from real work, rather than generic demos.
Leaders, sales, operations and engineering work on tasks from their own jobs, with shared usage rules.
We define what is allowed, restricted and prohibited before AI becomes part of daily work.
If the team does not leave the workshop with at least three workflows it can use straight away, you do not pay for the workshop.
The goal is faster, more orderly work on selected processes, with human review and clear boundaries.
How the offer works
Workflow audit
We map where the team loses time and select safe training examples, keeping sensitive data out of unapproved tools.
Role-based workshops
We adapt the workshops to leaders, business teams and engineering teams. Each group works on its own everyday tasks.
Hands-on sessions
The team creates drafts, summaries, research, technical tasks or review materials, following rules for checking and approval.
AI adoption report
We summarise what is useful, where caution is needed, which rules remain and which workflow is worth turning into an AI pilot.
Processes worth reviewing
How we check the result
Before the workshop, we define how the selected task is done today and where time is lost.
The team completes the same task using Claude, Claude Code or Codex, depending on the role and available environment.
We record where human review, source checks, anonymisation or a prohibition on use is needed.
We check which tasks the team can do independently. You keep templates, rules and recommendations for what to use immediately and what to test in a pilot first.
Boundaries before AI
- !We do not promise revenue, savings or productivity without measuring the real process.
- !We do not use sensitive data in personal or unapproved accounts.
- !We do not send AI outputs to customers, employees or partners without human review.
- !We do not introduce a tool before the rules: data, approvals, ownership and responsibility must be clear.
- !We do not treat Claude, Claude Code or Codex as substitutes for the team’s professional judgement.
Want your team to use AI on real work?
From €50/h · useful or free · reply within 24h
Frequently asked questions
An illustrative example: if five people each save two hours a week, that is around 40 hours a month. The actual benefit depends on tool use, output quality and the total cost of training and tools.
You do not pay for it. Every workshop aims to leave the team with at least three workflows it can use immediately, with templates and rules. If you do not receive that, the workshop is free.
Prices start at €50/h. We agree scope based on the number of people, team roles and the extent of hands-on sessions using your tasks.
No. Engineering teams can work on repositories and review preparation, while business teams work on emails, reports, research, CRM, SOPs and internal processes.
Not necessarily. We first check which tools and plans you are allowed to use. Workshops can use safe demos or anonymised examples.
You receive templates, usage rules, a list of safe workflows and a recommendation for a process worth developing further as an AI pilot or automation.