AI-Assisted Medical Diagnosis
AI-Assisted Medical Diagnosis
Year
Context
In France, 3 million people suffer from a rare disease, and 25% of patients wait between 5 to 15 years for a correct diagnosis. To address this "diagnostic odyssey," Sanofi, a pioneer in rare diseases, launched the AccelRare project, an AI solution "Made in France".
Context
In France, 3 million people suffer from a rare disease, and 25% of patients wait between 5 to 15 years for a correct diagnosis. To address this "diagnostic odyssey," Sanofi, a pioneer in rare diseases, launched the AccelRare project, an AI solution "Made in France".
Context
In France, 3 million people suffer from a rare disease, and 25% of patients wait between 5 to 15 years for a correct diagnosis. To address this "diagnostic odyssey," Sanofi, a pioneer in rare diseases, launched the AccelRare project, an AI solution "Made in France".
Context
In France, 3 million people suffer from a rare disease, and 25% of patients wait between 5 to 15 years for a correct diagnosis. To address this "diagnostic odyssey," Sanofi, a pioneer in rare diseases, launched the AccelRare project, an AI solution "Made in France".
Problem
The main challenge was to design an AI-powered tool to aid in the diagnosis of rare diseases, with a user experience simple and fast enough to be used by a general practitioner in under 10 minutes during a consultation.
Problem
The main challenge was to design an AI-powered tool to aid in the diagnosis of rare diseases, with a user experience simple and fast enough to be used by a general practitioner in under 10 minutes during a consultation.
Problem
The main challenge was to design an AI-powered tool to aid in the diagnosis of rare diseases, with a user experience simple and fast enough to be used by a general practitioner in under 10 minutes during a consultation.
Problem
The main challenge was to design an AI-powered tool to aid in the diagnosis of rare diseases, with a user experience simple and fast enough to be used by a general practitioner in under 10 minutes during a consultation.
Client
Sanofi
Industry
Health
Service
UI/UX Design
Web Development
Accessibility
Approach
The project was based on an existing AI algorithm integrated into a technical "Proof of Concept" with poor ergonomics. Our approach was therefore to lead a complete redesign of the user experience, focusing on building a bridge of trust between the physician and the technology, and transforming a technical tool into a simple and responsible clinical experience.
Approach
The project was based on an existing AI algorithm integrated into a technical "Proof of Concept" with poor ergonomics. Our approach was therefore to lead a complete redesign of the user experience, focusing on building a bridge of trust between the physician and the technology, and transforming a technical tool into a simple and responsible clinical experience.
Approach
The project was based on an existing AI algorithm integrated into a technical "Proof of Concept" with poor ergonomics. Our approach was therefore to lead a complete redesign of the user experience, focusing on building a bridge of trust between the physician and the technology, and transforming a technical tool into a simple and responsible clinical experience.
Approach
The project was based on an existing AI algorithm integrated into a technical "Proof of Concept" with poor ergonomics. Our approach was therefore to lead a complete redesign of the user experience, focusing on building a bridge of trust between the physician and the technology, and transforming a technical tool into a simple and responsible clinical experience.
UX Research
My role was to translate the power of the AI algorithm into an intuitive and reliable interface. This included:
Auditing the existing technical POC to identify key friction points.
Designing the wireframes to simplify the complex medical journey.
Creating the responsive UI (desktop & mobile) and a styleguide to ensure consistency.
Anticipating Accessibility Standards (WCAG) from the project's outset, which proved critical as it prevented a costly technical rework when digital accessibility later became a legal requirement.
UX Research
My role was to translate the power of the AI algorithm into an intuitive and reliable interface. This included:
Auditing the existing technical POC to identify key friction points.
Designing the wireframes to simplify the complex medical journey.
Creating the responsive UI (desktop & mobile) and a styleguide to ensure consistency.
Anticipating Accessibility Standards (WCAG) from the project's outset, which proved critical as it prevented a costly technical rework when digital accessibility later became a legal requirement.
UX Research
My role was to translate the power of the AI algorithm into an intuitive and reliable interface. This included:
Auditing the existing technical POC to identify key friction points.
Designing the wireframes to simplify the complex medical journey.
Creating the responsive UI (desktop & mobile) and a styleguide to ensure consistency.
Anticipating Accessibility Standards (WCAG) from the project's outset, which proved critical as it prevented a costly technical rework when digital accessibility later became a legal requirement.
UX Research
My role was to translate the power of the AI algorithm into an intuitive and reliable interface. This included:
Auditing the existing technical POC to identify key friction points.
Designing the wireframes to simplify the complex medical journey.
Creating the responsive UI (desktop & mobile) and a styleguide to ensure consistency.
Anticipating Accessibility Standards (WCAG) from the project's outset, which proved critical as it prevented a costly technical rework when digital accessibility later became a legal requirement.
Challenges
Designing this tool raised several complex challenges:
Designing for Trust: The main challenge was to make the suggestions of a "black box" algorithm transparent and credible for a healthcare professional, without ever replacing their medical judgment.
Simplifying Complexity: We had to translate complex probabilistic diagnostic data into a clear, quickly readable, and unambiguous interface.
Managing Responsibility: The interface had to encourage a positive "culture of doubt," presenting the results as leads to explore rather than absolute truths—a major ethical issue.
Challenges
Designing this tool raised several complex challenges:
Designing for Trust: The main challenge was to make the suggestions of a "black box" algorithm transparent and credible for a healthcare professional, without ever replacing their medical judgment.
Simplifying Complexity: We had to translate complex probabilistic diagnostic data into a clear, quickly readable, and unambiguous interface.
Managing Responsibility: The interface had to encourage a positive "culture of doubt," presenting the results as leads to explore rather than absolute truths—a major ethical issue.
Challenges
Designing this tool raised several complex challenges:
Designing for Trust: The main challenge was to make the suggestions of a "black box" algorithm transparent and credible for a healthcare professional, without ever replacing their medical judgment.
Simplifying Complexity: We had to translate complex probabilistic diagnostic data into a clear, quickly readable, and unambiguous interface.
Managing Responsibility: The interface had to encourage a positive "culture of doubt," presenting the results as leads to explore rather than absolute truths—a major ethical issue.
Challenges
Designing this tool raised several complex challenges:
Designing for Trust: The main challenge was to make the suggestions of a "black box" algorithm transparent and credible for a healthcare professional, without ever replacing their medical judgment.
Simplifying Complexity: We had to translate complex probabilistic diagnostic data into a clear, quickly readable, and unambiguous interface.
Managing Responsibility: The interface had to encourage a positive "culture of doubt," presenting the results as leads to explore rather than absolute truths—a major ethical issue.
Impact
The project's success validated our design approach by demonstrating the power of Human-AI collaboration. The results showed that while either the doctor alone or the AI alone had limited performance, their combination was spectacular.
The true achievement was proving that design could create synergy: by using the tool, the physician's diagnostic performance increased by over 40 percentage points. Our work didn't just make the AI useful; it augmented the practitioner's own abilities. The product was then successfully launched, with an ongoing deployment in 6 countries (including France, Mexico, and South Africa) and 21 in the future.
Impact
The project's success validated our design approach by demonstrating the power of Human-AI collaboration. The results showed that while either the doctor alone or the AI alone had limited performance, their combination was spectacular.
The true achievement was proving that design could create synergy: by using the tool, the physician's diagnostic performance increased by over 40 percentage points. Our work didn't just make the AI useful; it augmented the practitioner's own abilities. The product was then successfully launched, with an ongoing deployment in 6 countries (including France, Mexico, and South Africa) and 21 in the future.
Impact
The project's success validated our design approach by demonstrating the power of Human-AI collaboration. The results showed that while either the doctor alone or the AI alone had limited performance, their combination was spectacular.
The true achievement was proving that design could create synergy: by using the tool, the physician's diagnostic performance increased by over 40 percentage points. Our work didn't just make the AI useful; it augmented the practitioner's own abilities. The product was then successfully launched, with an ongoing deployment in 6 countries (including France, Mexico, and South Africa) and 21 in the future.
Impact
The project's success validated our design approach by demonstrating the power of Human-AI collaboration. The results showed that while either the doctor alone or the AI alone had limited performance, their combination was spectacular.
The true achievement was proving that design could create synergy: by using the tool, the physician's diagnostic performance increased by over 40 percentage points. Our work didn't just make the AI useful; it augmented the practitioner's own abilities. The product was then successfully launched, with an ongoing deployment in 6 countries (including France, Mexico, and South Africa) and 21 in the future.
Tools
🎨 Sketch then Figma for UI design
🧠 Miro for workshops and user research
⚙️ Marvel for interactive prototyping
Tools
🎨 Sketch then Figma for UI design
🧠 Miro for workshops and user research
⚙️ Marvel for interactive prototyping
Tools
🎨 Sketch then Figma for UI design
🧠 Miro for workshops and user research
⚙️ Marvel for interactive prototyping
Tools
🎨 Sketch then Figma for UI design
🧠 Miro for workshops and user research
⚙️ Marvel for interactive prototyping
PORTFOLIO
CHECK OUT SOME MORE
CHECK OUT SOME MORE
CHECK OUT SOME MORE
CHECK OUT SOME MORE
Spotlight UI : Puma Ignite
Spotlight UI : Puma Ignite
Spotlight UI : Puma Ignite
Art Direction
UI design
Spotlight UI : Puma Ignite
Spotlight UI : Puma Ignite
Spotlight UI : Puma Ignite
Art Direction
UI design
Spotlight UI : Puma Ignite
Spotlight UI : Puma Ignite
Spotlight UI : Puma Ignite
Art Direction
UI design
Spotlight UI : Puma Ignite
Spotlight UI : Puma Ignite
Spotlight UI : Puma Ignite
Art Direction
UI design
Banking API Portal
Banking API Portal
Banking API Portal
UI / UX Design
Web Development
Digital Design
Banking API Portal
Banking API Portal
Banking API Portal
UI / UX Design
Web Development
Digital Design
Banking API Portal
Banking API Portal
Banking API Portal
UI / UX Design
Web Development
Digital Design
Banking API Portal
Banking API Portal
Banking API Portal
UI / UX Design
Web Development
Digital Design
Let’s talk.
I’m always open to meaningful conversations around product, strategy or complex UX challenges.