AI applied to company professions — training

A tangible deliverable

Participants leave with tools, not notes

After the three days, each participant leaves with working tools they built themselves — configured business assistants, document templates, a prompt library, analysis grids — and a measure of the time they save. All with freely available AI tools: no subscription is needed to apply what is taught.

21 hover 3 days

70 %hands-on

6workshops

0 DApaid tools

Title	Artificial intelligence applied to company professions
Duration	3 days — 21 hours, in 6 half-days of 3 h 30
Format	In person, workshop format: about 30 % structuring input, 70 % hands-on production
Audience	Multidisciplinary audience: management, finance, HR, sales, marketing, technical, administration, quality
Group size	10 to 17 participants — beyond that, splitting the group is recommended
Prerequisites	No technical knowledge. Everyday use of a computer, the web and office software
Tools	Exclusively freely available AI tools
Languages	French and Arabic — materials, workshops and deliverables
Customisation	Content rebuilt from the sector, professions and real documents of the company

Standard programme: its structure is stable, its content is rebuilt for each company and finalised after the scoping interview.

What participants produce

Each workshop ends with a deliverable, built by the participant on their own workstation from their own files. They are working tools usable as soon as they are back at the office.

🤖

Configured business assistant

An assistant that knows the company’s vocabulary, procedures and tone, to whom a task can be entrusted without re-explaining everything.

📚

Team prompt library

The wordings that work are written once, named and shared: colleagues no longer have to rediscover them.

📝

Document templates

Report, minutes, memo, letter, specifications, product sheet: a constant structure, faster writing.

📊

File analysis grid

An Excel file queried in plain language, a summary and a recommendation produced in minutes.

🎨

Content kit

Texts and visuals for a campaign, internal communication or a presentation, produced without an agency.

📋

Written operating procedure

A procedure or instruction finally written down, because the writing effort is no longer an obstacle.

✅

Checklist and control grid

Systematic checking where everyone used to rely on memory.

🗓️

90-day action plan

Three priority uses, one tool, one deadline and one indicator per participant.

Three effects, measured by participants themselves

Productivity — the time spent on recurring writing and formatting tasks is estimated before, then after. Each participant quantifies their own gain.
Quality — documents gain structure and consistency: a common template, systematic proofreading, fewer omissions and fewer returns.
Deadlines — deliverables that were waiting to be written (reports, procedures, minutes, replies) stop being bottlenecks.

The journey

Three days to understand, produce, deploy

DAY 1

Understand and frame

Choose the right tool, talk to it effectively, and build your first business assistant. Confidentiality rules are set from the very first half-day.

Workshops 1 and 2

DAY 2

Produce and make reliable

Produce professional documents, content and visuals. Check, correct, structure: quality before volume.

Workshops 3 and 4

DAY 3

Analyse and deploy

Make your data speak, automate the repetitive, and settle a measurable action plan with a team usage charter.

Workshops 5 and 6

Detailed programme — 6 half-days

Day 1 · Session 1 — Framing generative AI in your profession

What generative AI can do, what it does poorly, what it invents: hallucination demonstrated live on a case from the sector.
Comparative overview of free assistants: strengths, limits, which tool for which task.
Confidentiality and data sovereignty: what must never go into a prompt.
Installation and getting started: accounts, mobile app, workstation.

Workshop 1 · Task mapping — in professional sub-groups, participants list their real tasks and sort them: automatable, assisted, non-delegable. Deliverable: a prioritisation matrix specific to the company.

Day 1 · Session 2 — The prompt, skill no. 1

Anatomy of an effective prompt: role, context, task, constraints, expected format, example.
Techniques that change the result: giving an example, step-by-step reasoning, asking for critique and rewrite, imposing a template.
Creating an assistant that knows the company: vocabulary, procedures, tone, prohibitions.
Building a shared prompt library: naming, versioning, circulation between departments.

Workshop 2 · The business assistant — everyone builds and tests their assistant, then writes five reusable prompts from their own job. Deliverable: configured assistant + first version of the shared library.

Day 2 · Session 3 — Writing and structuring your documents

Assisted professional writing: memo, report, minutes, letter, procedure, proposal, reply.
Building a reusable template rather than an isolated text: the difference between saving one hour and saving a hundred.
Adapting the register to the recipient: management, colleague, client, administration, foreign partner.
Translation and adaptation French / Arabic / English: what AI does well, what requires human proofreading.
Making it reliable: checking facts, figures and references. What is never delegated.

Workshop 3 · The team template — each sub-group turns a recurring company document into an assisted template, tested on two real cases. Deliverable: documented template, ready to circulate.

Day 2 · Session 4 — Creating visuals and presentations

Free image generators: realistic uses, limits, usage rights.
The visual prompt: subject, style, framing, lighting, format — and the art of variation.
Building a presentation: structuring the message, producing visuals, keeping graphic consistency.
The professional red line: never present a generated image as a real one. Rules on disclosure, rights, legal and reputational risk.

Workshop 4 · The complete deck — end-to-end production of a communication or presentation material: structure, texts, visuals. Deliverable: exportable material.

Day 3 · Session 5 — Making your data speak

Using an Excel or CSV file without being an analyst: import, cleaning, plain-language questions, summary tables.
Analysing the company’s business data: activity, costs, sales, incidents, headcount, monitoring indicators.
Analysing qualitative material: comments, complaints, field feedback, verbatims — recurring themes and weak signals.
Augmented research and monitoring: market, competition, regulation, tenders.
Reporting: turning an analysis into a clear message and a recommendation for management.

Workshop 5 · From data to decision — from an (anonymised) company dataset, each sub-group produces a commented analysis and a one-page recommendation. Deliverable: summary note.

Day 3 · Session 6 — Automating, governing, deploying

Automating the repetitive: standard replies, minutes, planning, translation, scheduled monitoring.
Introduction to specialised assistants and automated chains: the principle, realistic cases, and what must never be automated.
Governance: internal usage charter, data protection, systematic checking, disclosure of generated content, responsibility for errors.
Measuring: building a simple indicator for time saved, quality and deadline compliance.
Deploying: spreading these practices to the rest of the team without new training.

Workshop 6 · 90-day action plan — each participant formalises three priority uses, the chosen tool, the deadline and the indicator. The group jointly drafts the company’s usage charter. Final presentation, evaluation and handover of deliverables.

Tailor-made

Customised to your sector

AI training that stays generic changes no practice. This programme is therefore rebuilt before each session from the company’s reality: the six sessions and six workshops do not change, but the examples, documents handled, datasets analysed, vocabulary and prompts are yours.

Step	What is done	When
1. Positioning questionnaire	Sent to each participant: role, time-consuming tasks, digital comfort, expectations. It reveals the group’s real level and everyday irritants.	D-15
2. Scoping interview	45-minute exchange with the training manager and a business manager: sector issues, regulatory constraints, ongoing projects.	D-10
3. Collection of materials	Real documents and data: report templates, procedures, catalogues, anonymised file extracts. They become the substance of the workshops.	D-7

Examples of sector adaptation

Sector	Cases worked on in workshops
Industry and production	Operating procedures, instruction sheets, non-conformity reports, machine-downtime analysis, product data sheets
Banking, finance, insurance, leasing	Analysis notes, case summaries, regulatory letters, portfolio analysis, sales pitches
Retail and trade	Product sheets, network pitches, sales analysis by range and point of sale, promotional content
Services and consulting	Commercial proposals, mission reports, client deliverables, sector monitoring
Health and pharmaceuticals	Internal documentation, regulatory summaries, training materials, analysis of field feedback
Administration and institutions	Briefing notes, activity reports, administrative correspondence, use of public data
Energy and environment	CSR and extra-financial reports, awareness campaigns, technical summaries, performance indicators
Tourism and hospitality	Descriptions and programmes, replies to customer reviews, multilingual content, seasonality analysis
Construction and real estate	Specifications, technical memoranda, tender replies, site monitoring

This list is not exhaustive.

Multidisciplinary audience

A mixed group, a constraint turned into an asset

Different roles, seniority levels and digital comfort in the same room: the format prevents the group from splitting between those who are bored and those who drop out.

Common core — same notions, same vocabulary, same confidentiality rules: a shared culture between departments.
Workshops by professional affinity — everyone progresses on their own documents.
Mixed-level pairs — a confident participant is paired with a more hesitant one: the level rises for all.
Cross presentation — each sub-group presents its production: uses circulate between departments.
Challenge levels — a mandatory base and an extension for the fastest: nobody waits, nobody drops out.

👥

General management

A strategic reading: what AI changes in the company’s model, what must be arbitrated, what must be framed.

👥

Finance and accounting

File analysis, case summaries, writing of notes and reports.

👥

Human resources

Job descriptions, job offers, assisted shortlisting, training materials, minutes.

👥

Sales and marketing

Content, sales pitches, visual creation, analysis of sales and customer feedback.

👥

Technical and production

Operating procedures, documentation, incident reports, technical research.

👥

Administration and secretarial

Correspondence, meeting minutes, formatting, translation.

👥

Quality, HSE, CSR

Procedures, audit grids, compliance reports and extra-financial reports.

How a half-day unfolds

The training is not a lecture: a 3 h 30 half-day alternates short listening and production.

Sequence	Duration	What happens
Opening	15 min	Review of what has been tried since the previous session; difficulties met, successes shared.
Structuring input	30 min	The notion of the day, introduced through a business case from the company’s sector. Never more than 30 minutes at a stretch.
Commented demonstration	25 min	The trainer performs the task live on screen, deliberately makes mistakes, corrects, explains why.
Guided reproduction	35 min	Everyone repeats the operation on their own workstation, on their own document, with individual support.
Break	15 min
Sub-group workshop	60 min	Autonomous production by professional affinity: this is where the session’s deliverable is born.
Presentation and critique	30 min	Each sub-group presents; the group critiques; the trainer consolidates and corrects.

Interactivity levers

Timed challenges — produce a deliverable in a set time, then compare the group’s approaches.
The “weakest prompt” — a poor result is projected and the group looks for what was missing in the request.
Error hunt — a generated text containing inaccuracies is handed out; participants must spot them. This is how the verification reflex is built.
Interactive quizzes — short anchoring points at the end of a session, on smartphones, with commented correction.
Cross presentation — each sub-group explains its approach to the other departments.

Method, deliverables and evaluation

About 30 % structuring input, 70 % hands-on production; no lecture-style session longer than 30 minutes.
Work exclusively on the company’s real materials and data, shared in advance.
No code, no technical jargon: each notion is introduced through a business case.
Progression in three stages: the trainer shows, the participant reproduces, the sub-group produces alone.

Deliverables handed over

The full training material, available offline.
The company’s prompt library and the configured business assistants.
Workshop outputs: prioritisation matrix, templates, communication material, analysis note.
The draft AI usage charter, to be validated by management.
Each participant’s 90-day action plan and a memo sheet on tools.

Evaluation

Beforehand — positioning questionnaire and scoping interview.
Continuous — evaluation by production: each workshop produces an observable, commented deliverable.
At the end — knowledge-validation quiz and skills self-assessment grid.
Immediately after — satisfaction questionnaire sent to the client.
Later (option) — 30-day follow-up on progress of action plans and observed gains.

Why free tools, and why it is enough

The free versions of the major AI assistants cover almost all of a department’s needs. Each participant can apply immediately, the company measures real value before committing a budget, and the teaching stays independent: no paid solution is recommended unless requested, and the trainer has no commercial agreement with any vendor. If the company already holds licences, they are integrated into the workshops without being a prerequisite.

Artificial intelligence applied to company professions