AI foundations for decision-makers
What agents can and cannot do, use-case selection, data boundaries, vendor questions, human review, and an adoption plan.
Private AI training · Portugal and remote
Train your team on a real workflow, not a parade of AI tools.
I run live, private AI training for small engineering, operations, marketing, and leadership teams. The curriculum follows the work your people already own, and each longer cohort ends with a prototype, evaluation method, or operating playbook they can keep using.
Direct answer
A working programme for one team, adapted to its actual responsibilities.
Practical AI team training should change how a group chooses use cases, handles company context, checks output, and decides where human review belongs. I offer focused workshops and private cohorts from two to six weeks. We use representative work from the team, build together in live sessions, and document what can be repeated after the training. This is not open enrolment, a vendor certification, or a catalogue of prompt tricks.
Related work
One technical problem often crosses several systems.
The capabilities page shows the other ways I can help and the case studies behind them.
When this helps
Useful when individual experiments need to become a shared, reviewable way of working.
- People use AI separately, but the team has no shared quality or data-handling method.
- A workflow looks automatable, yet nobody has tested the exceptions or operating risk.
- Engineers can prototype agents but need a clearer approach to evaluation, permissions, and release.
- Leaders need enough technical understanding to challenge proposals and set responsible boundaries.
- Marketing or publishing teams need source-aware AI workflows rather than faster generic content.
What the work covers
Four tracks cover the roles I can teach from hands-on work.
We choose one primary track and adapt the examples, tools, pace, and capstone to the team. A mixed cohort can combine foundations with one practical lab.
Workflow automation lab
Map a recurring process, separate deterministic and AI-assisted steps, connect tools, test exceptions, and assign an owner.
AI agent engineering sprint
Define the agent contract, context, tools, permissions, evaluations, logs, fallback, and a controlled release path.
AI search and publishing lab
Build evidence-led research, source tracking, structured publishing, review, and measurement into a repeatable workflow.
Ways to train your team
Invest in one team and one operating outcome—not a seat licence.
Each level covers a private group of up to ten people and tailoring around one representative workflow. Choose a briefing for alignment, a workshop for one method, or a cohort when practice and a capstone must change how the team works.
AI leadership briefing
€950fixed
A shared vocabulary, use-case map, risk questions, and a responsible next step for decision-makers.
- Preparation call
- Live private briefing
- Decision and risk summary
Practical AI workshop
€1,800fixed
A focused live session that teaches and applies one bounded method to the team’s work.
- Tailored examples
- Three-and-a-half to four live hours
- Templates and next-action plan
Full-day AI workshop
€2,900fixed
A deeper working day with more practice, review, and an agreed operating method.
- Tailored preparation
- Approximately six live hours
- Worked example and handover notes
AI workflow cohort
From€5,500programme
Practice between sessions and complete a capstone based on a real team workflow.
- Live sessions and office hours
- Prototype or operating playbook
- Evaluation cases and 30-day plan
AI agent engineering cohort
From€12,500programme
A controlled agent sprint covering the system around the model and a production-minded handover.
- Agent contract and architecture
- Evaluation, permissions, logs, and fallback
- Capstone review and operating handover
Good to know
The fee covers the learning programme; production implementation stays separate.
- Workshop bookings are 50% due on confirmation and 50% before delivery.
- Figures exclude VAT, software licences, API usage, travel, and accommodation.
- Remote delivery is the default; on-site work is scoped separately.
- Custom production implementation, certification, and unlimited follow-up are not included unless stated in the proposal.
EURAvailable worldwide. All fees are quoted and invoiced in EUR. Your bank or payment provider may convert the amount into your local currency at its current rate.
Compare the options
Choose the format by the behaviour that needs to change.
| Format | Best when | Result |
|---|---|---|
| Leadership briefing | Decision-makers need a common vocabulary and responsible next step. | A use-case map, risk questions, and adoption priorities. |
| Hands-on workshop | A small group needs to learn and test one bounded method together. | A worked example, templates, and a next-action plan. |
| Private cohort | The team needs practice between sessions and a capstone on real work. | A prototype or playbook, evaluation cases, and handover. |
| Build plus enablement | A production delivery must become understandable and operable by the internal team. | Working software, guided practice, operating notes, and ownership. |
How the work runs
Diagnose the gap, tailor the practice, build live, and leave the method behind.
Choose the outcome
Identify the team, current AI use, target workflow, constraints, and the behaviour that should change.
Tailor the programme
Select the track, representative cases, tools, schedule, preparation, and capstone boundary.
Work in live sessions
Teach the concepts briefly, apply them to the team’s work, review failures, and improve the method together.
Hand over
Deliver the prototype or playbook, evaluation cases, responsibilities, open risks, and the next 30 days of practice.
Relevant work
The teaching model comes from engineering work and project-based education.
I trained at 42 Paris, worked on the 42 Network student intranet, and joined the founding team of 42 Lisboa in IT and pedagogy. I now build AI workflows, agents, products, and publishing systems. That background supports practical, peer-led learning tied to working software; it is not a claim of formal AI certification.
FAQ
Questions before planning private AI training.
How many people can join?
The practical range is three to ten participants. A leadership briefing can include a larger group, but hands-on work needs enough time for each person and one accountable team owner.
Do participants need to code?
Only the agent engineering track assumes software development experience. Foundations, workflow automation, and publishing tracks can use no-code tools or the systems already available to the team.
Can we use our own workflow and data?
Yes, and the programme is stronger when we can. We agree on access, privacy, redaction, test data, and what must remain outside the sessions before the work begins.
Is this available on-site?
Remote delivery is the default. On-site workshops in Portugal, and selected European locations, can be scoped when being in the room materially improves the work.
What does the team receive?
Every engagement includes the tailored session material and a practical record of the work. Longer cohorts also include a prototype or playbook, evaluation cases, open risks, and a 30-day continuation plan.
Do you provide certification?
No. This is applied professional training tied to the team’s work, not an accredited qualification or vendor certification.
What does private AI training start at?
A 90-minute leadership briefing is €950, a half-day private workshop is €1,800, and a full day is €2,900. Private 2–4 week cohorts start at €5,500; a six-week agent engineering cohort starts at €12,500. Each format covers one private group of up to ten people and excludes VAT, travel, software, and API usage.
Start with the team and one workflow
Tell me who needs to learn, what they already use, and what should work differently afterwards.
I will recommend a briefing, workshop, private cohort, build-plus-enablement engagement, or a different specialist if the requested track falls outside my evidence.
Plan the training