← All projects

Serious game · business case exercise

OrIA

A serious game that demystifies AI for Orange managers who distrust it or don’t know how to use it: each level gives you manager tasks and a time budget, and you pick the right software… or the right prompt.

OrIA title screen: the game logo and the Orange logo, with a “Play” button.

Fact sheet

Engine
Unity
Year
2025
Duration
About 1 month
Team
5 people
Platform
Browser (WebGL, Unity Play)

Tools & technologies

  • Unity 6
  • C#
  • ScriptableObjects
  • WebGL
  • Figma

My role

My part

Team project : 5 people. This list only covers my part.

  • Helped lay the foundations of the game design as a team: the game’s structure and how it plays
  • Interface mock-up in Figma
  • ScriptableObject data architecture: days, tasks, software and prompts
  • Most of the game manager: levels, days, menus
  • Much of the front-end integration of the interface, with help from the team: task selection, software and Mistral AI windows, confirmations, evaluation, level navigation
  • Web build (WebGL)

The brief

As part of a course, a senior staff member at Orange came to us with an observation: in their company, some employees, especially older ones, feel left behind by AI. Some are afraid of it, others don’t know how to use it. The exercise: design a serious game to demystify it. It was not a commission from the company.

How it plays

The game is played with the mouse, level by level. Each level gives you a number of tasks and a time budget to manage. Picking a task, such as “prepare a monthly roadmap”, opens an interface that simulates a Windows desktop with four programs: Mistral AI and three tools that may or may not suit the task (here Excel, PowerPoint and Trello).

A poorly suited tool completes the task but costs more time. In Mistral AI you choose between four prompts: a bad prompt wastes time without completing the task. Every choice gets graded feedback: poor, good, great or excellent.

The message

Not everything is a job for AI: handling a team conflict, for example, is better done through Outlook or Teams. The tasks and software are based on real cases from Orange managers.

How we worked

An agile process, with a weekly progress review with the Orange staff member.

Good to know

AI disclosure

AI-assisted code. The answers shown in the Mistral AI window are real answers generated by Mistral AI from the game’s prompts, then added to the game.