Saima Ben Hadj

As VP of Data Science, I led a multinational team of AI scientists, engineers, and a project manager, all united by a mission to outsmart cancer with AI. We focus on developing AI tools for faster, more accurate diagnoses, especially in breast cancer, resulting in CE IVD marked solutions. Beyond the algorithms, I bridge the gap between tech, medical experts and business teams to ensure our AI solutions are as impactful in the real world as they are on paper. It’s about leading with vision, a bit of grit, and a commitment to making a real difference in healthcare.

As Head of AI, I led the creation and shaping of the AI department, recruiting and guiding a team of skilled data scientists. My role involves defining and executing long-term strategies to develop robust, reliable AI solutions in digital pathology, addressing the challenges posed by the variability of cancer biology and image acquisition systems. I ensure our projects are not only technically sound but also cost-effective and ethically aligned.

As a Computer Vision Engineer, I developed the company’s first machine learning tools. I specialized in creating automated solutions for histopathology image analysis, focusing on segmentation, registration, and stain unmixing using advanced machine learning and deep learning techniques. I also played a key role in AI-focused education and client support, boosting their engagement and understanding of our solutions.

In my Ph.D. research, I specialized in computer vision, focusing on mathematical inverse problems to improve the quality of 3D fluorescence microscopy images. I developed advanced convolution and deconvolution techniques, combined with functional optimization, to reduce image blur and noise. My work involved tackling complex, high-dimensional non-convex optimization problems and creating algorithms for regularized cost functions. This research resulted in seven first-author publications in prestigious journals and conferences, including IEEE and JIMIV. Additionally, I contributed to academia by lecturing at the University of Nice.

I focused on automating object detection in aerial and satellite imagery, identifying objects such as boats, flamingos, trees, and buildings to support aerospace studies led by CNES (National Centre for Space Studies). I applied hyperparameter estimation algorithms with Markov Random Fields to improve detection accuracy. My work resulted in the publication of two research articles and earned me the Best Paper Award at SFPT’11.