Marta Hovanesian
In-company training
Doel van de training
Teams leren veilig en verantwoord AI gebruiken, met duidelijke kaders: wat mag wel, wat niet, en hoe voorkom je dat je onbewust data of IP “weggeeft”.
Leerdoelen (na 90 min kunnen deelnemers)
-
Uitleggen wat AI-tools doen met input en waarom dat risico’s geeft.
-
Herkennen welke data nooit in publieke AI-tools mag.
-
Werken met een simpele beslisboom: mag dit erin, ja/nee?
-
Output controleren: hallucinaties, auteursrecht, vertrouwelijkheid.
-
Meenemen wat er minimaal nodig is: AI-beleid + awareness + toolingkeuze.
Opzet (agenda)
0–10 min | Kick-off
-
Waarom AI wel, maar met regie
-
Mini-poll: “Wie gebruikt AI al in werk?”
10–25 min | Hoe AI werkt (zonder technisch gedoe)
-
Input → model → output (waar gaat het mis?)
-
Voorbeelden van “goed antwoord, verkeerde bron”
25–45 min | De 4 grootste risico’s (herkenbaar en concreet)
-
Vertrouwelijke info lekt (klantdata, strategie, contracten)
-
IP/auteursrecht (AI-gegenereerde content, images, code)
-
Onjuiste output (hallucinaties) → fout advies/communicatie
-
Compliance & reputatie (transparantie, logging, accountability)
45–70 min | Praktijk: 3 cases (groepjes)
-
Case 1: Marketing plakt klantcase in AI voor social captions
-
Case 2: HR laat AI een verzuimtemplate maken met echte details
-
Case 3: Sales gebruikt AI om een voorstel te maken met interne prijsafspraken
➡️ Per case: wat is het risico, wat is alternatief, welke regel hoort erbij?
70–85 min | AI-regels die je morgen kunt invoeren
-
“Stoplight model”: Groen / Oranje / Rood use-cases
-
Minimum set: beleid, training, toegestane tools, reviewproces
-
Wanneer escaleren naar legal/security
85–90 min | Afsluiting
-
5-min checklist + Q&A
-
Volgende stap: beleid-scan of team-specifieke verdieping
Deliverables (wat de klant krijgt)
-
1-pager AI Do’s & Don’ts (per afdeling)
-
Stoplight use-case matrix (groen/oranje/rood)
-
Beslisboom: “Mag dit in AI?”
-
Template: AI policy outline (basis voor hun eigen beleid)
Prijs (pilot-voorstel)
-
Pilot 90 min: €750 – €1.250 (afhankelijk van voorbereiding)
-
Optioneel: voorbereidende intake (30 min) + analyse van hun tools/workflows
Who this training is for
This is designed for organisations where AI is already in daily work: HR, marketing, sales, product and innovation teams, operations, IT and leadership. It’s equally valuable for fast-moving scale-ups and for established organisations that want consistent governance.
If your company wants the productivity upside of AI and the trust that comes from doing it responsibly, this training is the fastest way to get there.
Who this training is for
Marvenq delivers this training as a 60–90 minute high-impact session (ideal for a first rollout) or as a half-day workshop with deeper department-specific cases and implementation planning. It can also be delivered as a boardroom session focused on governance, accountability and decision-making across Cyber / AI / Data.
After the session, you don’t just have a “better understanding.” You have structure. Teams leave with guardrails they can apply immediately: a simple decision framework, a stoplight model (what’s safe, what’s sensitive, what’s not allowed), and a clear set of AI do’s and don’ts that fit real workflows. Leadership gets a clearer view on ownership and escalation: who decides what tools are permitted, what needs review, and what must be controlled to reduce risk without killing productivity. And most importantly: your organisation moves from a reactive posture hoping nothing goes wrong to an intentional one: using AI responsibly, with confidence, at scale.
FAQs
How long is the training?
Most organisations start with 60–90 minutes. If you want hands-on cases and rollout planning, choose a half-day workshop.
Is this training technical?
No. It’s designed for teams, managers and leadership. Clear language, practical examples, real decisions.
Can you tailor it to specific departments?
Yes—HR, Marketing, Product/Tech, IT/Security, Finance, Sales, or mixed groups.
Do we receive a policy after the training?
You receive a Safe AI policy outline plus practical rules and frameworks. Marvenq can also support implementation if needed.