AI campaign generation lets you turn an existing course into a complete SPARTED campaign, with varied gameplays, images and a title.
The AI analyses your course content, suggests a pedagogical plan for you to review, then generates all the gameplays.
If you don't have a course yet and want to create one from scratch, check out the associated article first:
➡️
Comme pour les parcours, le processus se fait en deux temps avec une étape de validation au milieu. Vous gardez la main sur la structure avant que le contenu soit généré.
La différence avec la génération de parcours : ici vous partez d'un parcours déjà existant. L'IA le lit, le comprend, et construit une campagne cohérente à partir de ce contenu.
📋 Prérequis
Like with courses, the process happens in two stages with a validation step in between. You stay in control of the structure before any content is generated.
The difference from course generation: here you start from an existing course. The AI reads it, understands it, and builds a coherent campaign from that content.
💡 Tips : take care of your source course. The AI generates the campaign from your micro-knowledge content. The better written, structured and contextual the course, the more relevant the campaign will be. A course with micro-knowledges that are too short or vague will produce a lower-quality campaign.
campagne de qualité moindre.
⚙️ Les phases du processus
Phase 1 — Reading the course
You click "Generate campaign". The system reads your entire course and sends it to the AI microservice. The execution starts immediately.
Phase 2 — Pedagogical plan generation (AI)
The AI analyses the course content and proposes a pedagogical plan: an ordered list of gameplays, each associated with a type (MCQ, Swiping, Ordering, etc.) and a specific learning objective. It decides on the progression and variety of formats.
Phase 3 — You validate the plan
This is the most important step in the process. The pedagogical plan defines how many gameplays will be generated and with what objective.
What you see at this stage: a numbered list of gameplays, each showing the format type chosen by the AI and the concept or skill the gameplay is meant to address.
What you can do: change the gameplay type for any line (for example swap an MCQ for an Ordering), rewrite the learning objective to make it more precise, remove a gameplay that doesn't seem relevant, or reorder them.
💡 Tip: be specific about objectives. The learning objective for each line directly guides content generation. "Understand safety rules" will produce something generic. "Identify the 3 situations that require wearing PPE" will guide the AI much more effectively
⚠️ Validation is irreversible. Once the plan is confirmed, content generation starts. The structure can no longer be modified automatically through the AI.
The AI can generate five gameplay types. Here is what each one produces in practice.
MCQ: a question with several proposed answers, one or more of which are correct. The AI generates the question, the choices, and an explanation shown after answering.
Swiping cards: cards to swipe towards zones (left, right, up, down). The AI defines the zones, the cards, and the correct assignment for each. Ideal for rankings, associations, and sorting.
Ordering cards: items to put back in the right order, following a top-to-bottom logic. The AI generates the items to order and the instructions. Ideal for procedures, steps, and priorities.
Error text: two versions of the same text are shown, one correct and one containing a mistake. The player must identify the right one. Ideal for rules, precise wording, and compliance.
Open question: an open-ended question with a reference answer shown afterwards. The player writes their own response and compares it to the correction. Ideal for reflection, scenarios, and soft skills.
Phase 4 — Gameplay generation (AI)
The AI takes the validated plan and generates all the content: questions, answers, texts, images, campaign title and description. Everything happens in the background.
Phase 5 — Campaign available
The campaign is created as a draft in your back-office. It is ready to be configured, completed and published.
🏗️ What the campaign contains at the end
Once generated, your campaign contains the gameplays automatically distributed across sessions. The number of sessions depends on how many gameplays are in your plan: 8 gameplays give 2 sessions, 12 give 3, and so on.
The campaign is created with a title, description and cover image generated by the AI. It is configured by default in flexible mode, active Monday to Friday, starting on the current day.
💡 Tip: review the configuration before publishing. The default settings (dates, active days, number of sessions) work for many cases, but check them before going live. The title and description can be edited directly from the campaign p
Comments
0 comments
Article is closed for comments.