Context
CEA (French Alternative Energies and Atomic Energy Commission) is a French public government-funded research organisation in the areas of energy, defense and security, information technologies and health technologies.
It means that they onboard a lot of projects from research to production in those areas. On top of this, I worked with the team responsible for the Witsee project.
Witsee is a software that enables CEA researchers to analyze scanners through AI machine learning.
This innovative solution helps research teams interpret and process medical imaging data in an automated and precise manner.
I joined Witsee as a Lead Product Designer, with the mission to design and optimize the platform's user experience. I arrived during the implementation of a major new feature: scanner analysis through fluorescence technology.
This new feature represented a significant technical and UX challenge, requiring an intuitive interface to allow researchers to effectively leverage this advanced technology.
The team was composed of researchers, developers, data scientists, and a Product Manager, working in a demanding R&D environment.
After a quick kick-off, we identified several challenges to address: A constraint: CEA being a public actor with significant administrative friction, we had to develop this new feature within a minimum number of working days while maintaining required quality standards.
A goal: CEA planned to accelerate Witsee's deployment mid-2024 and take it from R&D to Production. To ensure that, they needed a robust and scalable interface allowing researchers to easily use fluorescence analysis across different types of scanners and use cases.
A subgoal: Ensure that the new feature integrates seamlessly into the existing Witsee ecosystem, maintaining UX consistency while introducing specific interactions for fluorescence analysis.
Audit
Building Strong Foundations for Witsee's Fluorescence Feature
A good product feature isn't just about adding functionality—it starts with strong foundations to support the entire user experience. So, we began this journey with 3 days dedicated to workshops and ground rules for implementing the fluorescence analysis feature.
Design Principles First: We needed to establish basic rules to guide our design choices and decisions. What values must we respect when designing for researchers? How should the fluorescence analysis interface integrate with existing workflows?
After a quick ideation phase, we clustered concepts and refined them. Since this feature would become a central part of researchers' daily work, our principles needed to shape the team's approach. We concluded with an alignment session to ensure everyone was on the same page.
Our design principles were: intuitive, precise, efficient, reliable
Consistent Design Language: For quick decisions, we agreed to base our thinking on Witsee's existing design system while extending it for fluorescence-specific needs. This led us to define interaction patterns and visual hierarchy specific to medical imaging analysis.
We established naming conventions for fluorescence analysis states and parameters, keeping them simple and scientific. Following the "if a researcher can understand it immediately, so can everyone" principle, we decided on a clear 5-level intensity scale for fluorescence data visualization, from minimal (1) to maximum (5) intensity.
User-Centered Tone: For a scientific product used by researchers and medical professionals, clarity and precision were more important than personality. We defined a communication approach based on being factual, supportive, and trustworthy—ensuring that every interface message would help researchers make confident decisions when analyzing fluorescence data.
This foundation work ensured that the new fluorescence feature would seamlessly integrate into Witsee's ecosystem while meeting the specific needs of medical imaging analysis.
Conclusion
Since it was an R&D project within a limited timeframe with a small team, there were no measured results except team satisfaction. I was truly happy to work with them! I had extensive experience in product design but it was the first time I worked on implementing such a technically complex feature in the medical field. And it was exciting. I discovered this particular open-mindedness of researchers, probably one of the best teams I've worked with.
The collaboration with data scientists and CEA researchers allowed me to understand the subtleties of fluorescence analysis and create an interface that truly respects their workflows. Seeing the feature come to life and being used by researchers for their analyses was extremely rewarding.
The Witsee team continued developing this feature after my involvement, and I was delighted to learn that it had been successfully integrated into the production version, allowing CEA researchers to fully leverage fluorescence technology in their scanner analyses.
