Home

Works

About

in

CV

Home

Works

About

in

CV

Home

Works

About

in

CV

When?

Q3 - 2025

Team size:

3 People

Duration:

2 weeks

Platform:

Website

Context

Mistral AI is a French leader in open-source AI, positioned as a key European competitor to OpenAI through its flagship chatbot, "Le Chat."

Growth

Its market value skyrocketed from $6.2 billion to nearly $14 billion by early 2025.

Status

Despite its performance, Le Chat struggles with user stickiness.

Growth

Its market value skyrocketed from $6.2 billion to nearly $14 billion by early 2025.

Status

Despite its performance, Le Chat struggles with user stickiness.

THE CHALLENGE

THE CHALLENGE

How might we gamify Le Chat to drive engagement and retention?

How did we work

1

2

3

4

Benchmark
& Ideation

Design

User Testing

Iteration

How did we work

1

2

3

4

Benchmark
& Ideation

Design

User Testing

Iteration

Phase 1 : Benchmark & Ideation

Exploring competitive and related service.

Our first step was to research and benchmark the applications that make the best use of gamification in order to better understand which approaches work best.

Our first step was to research and benchmark the applications that make the best use of gamification in order to better understand which approaches work best.

Duolingo

Duolingo

Gamification

Gamification

Personalization

Personalization

🔍 It's got a fun and simple design, and rewards like daily bonuses, badges, and points encourage learners to stay engaged and come back regularly.

Nike Run Club

Nike Run Club

Reward system

Reward system

🔍 Nike Run Club encourage their users to exercise daily by mixing a reward system and a community approach to boost engagement and community development.

Turn those ideas into concepts.

We came up with many ideas, some more realistic than others and keep the ones that reflect our desire to create this bond of trust between the product and the user.

What we kept

What we kept

01 - Educational challenges

At the heart of Mistral's operations is the delivery of information and support for creativity and knowledge-based work. We hope that the challenges will keep this same spirit, allowing users to learn while playing.

02 - Personalized assistant

We want to make this game more playful and enjoyable, making users feel that this experience is unique.

03 - Scoreboards

We didn't choose the leaderboard score feature because Mistral doesn't have a user social network.

04 - Interactive challenges

We didn't go for Interactive Challenges because they just don't fit with the Mistral brand.

What we didn’t keep

What we didn’t keep

What we didn’t keep

03 - Scoreboards

We didn't choose the leaderboard score feature because Mistral doesn't have a user social network.

04 - Interactive challenges

We didn't go for Interactive Challenges because they just don't fit with the Mistral brand.

Phase 2: Design

Phase 2: Design

Phase 2: Design

After the ideation phase, we turned our most promising hypotheses into prototypes.

Onboarding Page

A new feature into the sidebar called “Chassistant”.

A new feature into the sidebar called “Chassistant”.

And playful mascot encouraging users the user when they play the challenges.

And playful mascot encouraging users the user when they play the challenges.

Playing Page

These games target three main user types: creative writers, knowledge seekers, and programmers. Each game offers multiple difficulty levels to accommodate users of varying skill levels.

These games target three main user types: creative writers, knowledge seekers, and programmers. Each game offers multiple difficulty levels to accommodate users of varying skill levels.

These games aim to facilitate learning through interactive and challenging games.

These games aim to facilitate learning through interactive and challenging games.

Reward System

After playing the games, users can earn one badges.

After playing the games, users can earn one badges.

And after collecting 3 badges, users can get one trophy and unlock a new game.

And after collecting 3 badges, users can get one trophy and unlock a new game.

Reward of badge

Reward of trophy

Phase 3: User testing

To deliver a succesful solution and create a great use experience,
we perfomed a series of user tests.

To deliver a succesful solution and create a great use experience, we perfomed a series of user tests.

With users to achieve the following goals:

With users to achieve the following goals:

To validate the fluidity and clarity of the challenges.

To validate the fluidity and clarity of the challenges.

To validate that our games are easily understandable by users and that the game' flow is accessible and stimulating.

To validate that our games are easily understandable by users and that the game' flow is accessible and stimulating.

To assess the interest and understanding of Chassistant' role as a mascot and supporing character.

To assess the interest and understanding of Chassistant' role as a mascot and supporing character.

Test Parameters

  • Age range: 26-43 years old

  • Daily AI users

  • Age range: 26-43 years old

  • Daily AI users

45 min • Google Meet

45 min • Google Meet

2 men and 3 women

2 men and 3 women

What worked

4/5 panenlist

  1. The users found the games adapted well to their learning styles.

“le niveau 1, c’e n'était ni trop facile ni trop difficile — ça donne envie d'essayer le niveau suivant de "Content Craft".

5/5 panenlist

  1. Understanding the game mechanics.

“C'est sympa de chatter avec Chassistant en jouant au jeu.”

4/5 panenlist

  1. Chassistant Creates Friendly Atmosphere.

"J’aime bien le chat avec l’animation et cool on peut avoir des conversation"

What didn’t work

4/5 panenlist

  1. Some users did not understand how to start the game when facing the “Play” CTA and the chat field.

“Je comprends pas, car on peut toujour de lui parler et cliquer”

3/5 panenlist

  1. They are thinking that additional instructions are needed to better understand the game's progression

“On s’attend à trouver plus d’instructions, de mise en contexte.”

4/5 panenlist

  1. The badge system does not offer sufficient motivation to encourage progression.

“J'ai gagné un badge, mais je ne sais pas trop ce que ça change, à part un nouveau jeu”

Phase 4: Iteration

Phase 4: Iteration

After gaining these insights from user tests, we optimized our solution.

#1

Optimization of Onboarding page

Removed the sidebar block from the corner.

Removed the sidebar block from the corner.

Placed the mascot and call-to-action (CTA) directly on the homepage, and added a button that can be hidden.

Placed the mascot and call-to-action (CTA) directly on the homepage, and added a button that can be hidden.

#2

#2

Improving User Guidance

Improved the wording to better introduce users to the game system.

Improved the wording to better introduce users to the game system.

Enhanced UI for a clear experience.

Enhanced UI for a clear experience.

#3

#3

Badge Redesign Aligned with Mistral

Redesigned the badges to make them look more special and align with Mistral’s brand image.

Redesigned the badges to make them look more special and align with Mistral’s brand image.

Final Prototype

Through several improvements, we built our solution:

Ornikar

Choose

What didn’t work

4/5 panenlist

  1. Some users did not understand how to start the game when facing the “Play” CTA and the chat field.

“Je comprends pas, car on peut toujour de lui parler et cliquer”

3/5 panenlist

  1. They are thinking that additional instructions are needed to better understand the game's progression.

“On s’attend à trouver plus d’instructions, de mise en contexte.”

4/5 panenlist

  1. The badge system does not offer sufficient motivation to encourage progression.

“J'ai gagné un badge, mais je ne sais pas trop ce que ça change, à part un nouveau jeu”

Ornikar

Choose

Ornikar

Choose

Get in touch

Let's collaborate and create something great!

Let's collaborate and create something great!

in

Linkedin

CV

CV

Mail

Zhonglin CHEN - ©2026

Duration:

2 weeks

Platform:

Mobile

Team size:

3 People

When?

Q3 - 2025

How did we work

1

2

3

4

Benchmark
& Ideation

Design

User testing

Iteration

What worked

4/5 panenlist

  1. The users found the games adapted well to their learning styles.

“le niveau 1, c’e n'était ni trop facile ni trop difficile — ça donne envie d'essayer le niveau suivant de "Content Craft".

5/5 panenlist

  1. Understanding the game mechanics.

“C'est sympa de chatter avec Chassistant en jouant au jeu.”

4/5 panenlist

  1. Chassistant Creates Friendly Atmosphere.

"J’aime bien le chat avec l’animation et cool on peut avoir des conversation"

Ornikar

Choose

Final Prototype

Through several improvements, we built our solution: