Lyon, Auvergne-Rhône-Alpes, France
I am eager to play a major role in the development of 𝐫𝐞𝐚𝐥-𝐥𝐢𝐟𝐞 𝐩𝐫𝐨𝐛𝐥𝐞𝐦-𝐬𝐨𝐥𝐯𝐢𝐧𝐠 𝐀𝐈 𝐬𝐨𝐥𝐮𝐭𝐢𝐨𝐧𝐬 𝐟𝐨𝐫 𝐦𝐞𝐝𝐢𝐜𝐢𝐧𝐞. I worked on 𝐜𝐥𝐚𝐬𝐬 𝐈, 𝐈𝐈a & 𝐈𝐈𝐈 medical devices as well as on 𝐃𝐞𝐞𝐩 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 using 𝐌𝐑𝐈, 𝐂𝐓 and 𝐗-𝐫𝐚𝐲 images. I am co-author of papers on the MICCAI SASHIMI workshop and in the Journal of Arthroplasty, as well as of abstracts on several medical conferences. 𝐀𝐫𝐞𝐚𝐬 𝐨𝐟 𝐞𝐱𝐩𝐞𝐫𝐭𝐢𝐬𝐞 : - Computer Vision: Segmentation, Classification, Detection, Landmarks/Keypoints, Heatmaps - Python, Pytorch, Pytorch-Lightning, Keras - Medical Images, Pseudonymization, Anonymization - Model serving: FastAPI, Flask - DICOM: DICOM protocol, PynetDicom - Docker, MLOps - Unittests, Pytest, CI/CD - Interested by NLP (Transformers) and TypeScript+React 𝐒𝐤𝐢𝐥𝐥𝐬: - Strong communicator w/ stakeholders and colleagues of various backgrounds - Trilingual + Good level in italian - Autonomous, Curious, Self-motivated, Fast learner, Autodidact - Clean code: Library ownership, Git & GitHub processes, Testing, CI/CD Github: https://github.com/FlorianMF
Management of a team of 4 AWS, Infrastructure as Code (IaC), Terraform Animal imaging MRI, CT Product strategy & roadmap
3D and 2D segmentation of coronary arteries and pathologies (calcifications, stenoses, plaque, occlusions, etc.) on CT images (CCTA). Classification and regression of stenosis, ischemia (FFR) and calcium scores. Model Deployment: Serving, FastAPI, Flask, Docker, CI/CD, MLOps, Azure, Ansible, DICOM protocol Data Science Team Lead. Onboarding and training of newcomers. Management of Data Scientists. Link with the Regulatory Affairs & Quality Assurance (RAQA) department. Technical documentation, verification. Training of staff regarding SOPs (e.g. Design & Development). ISO Audits, Gap analyses, MDD, MDR. Elaboration of FDA 510k submission (clinical validation protocol, technical documentation, ...) Project Management. Responsible of Jira & Confluence (Atlassian suite). Definition of data pseudonymisation protocols (minimisation, removal, ...).
Responsible for data science code architecture : organization and maintenance of the code structure, consistency and documentation (proprietary Python libraries). Responsible for data science team work environment : arbitration, organization, maintenance and training of tools and frameworks. Responsible for data collection : arbitration, maintenance and optimization Responsible for the technical watch and continuous training of the data science team on best practices and technical innovations. Responsible for the preparation and maintenance of technical documentation (CIR, QARA, …) in collaboration with the software development and operations teams. Conception, optimization, validation and testing of artificial intelligence algorithms. Development of the research activity (scientific watch, scientific communications, ...). Participation in the strategy and implementation of data collection and annotation Participation in the implementation of the algorithms’ integration into the products and services. Coordination of hardware and software security management of proprietary and paid third party servers: arbitration, monitoring, updating, and training of new staff. Coordination between the Data Science and the Software Development teams
Development of AI-based products for healthcare Develop analysis schemes in direct collaboration with surgeons and physicians (stakeholders and domain experts) Making the bridge between the Data Science, software development and the clinical team. Data Collection, Pseudonymization, Management and Engineering Annotation task definition, management and reviewing Deep Learning Research and Development (Prototyping, training, testing, visualizing, deploying) Technological watch and knowledge sharing Model Deployment: Serving, Torchserve, FastAPI, Docker, MLFlow, CI/CD, MLOps, Cloud Instance Management Best practices: Git, Testing (Pytest, coverage), clean code, documentation, teaching of others about best practices Publications (journals and congresses/conferences) Documentation and Filing, Regulatory & CE compliance Strategy development System Administration and Server customization (Linux)
Receive Calls from members of the ADAC (biggest german car club), treat their cases, organize services (e.g. towering, hotels, replacement vehicles). B2B, B2C, client support, conflict resolution.
"Development and Evaluation of Methods for the Automated Analysis of Knee Cartilage in MRI data" Machine Learning in Python (Deep Learning, Computer Vision, Image Processing, Pytorch, Numpy, Scipy, Pandas, OpenCV, Scikit-Learn, ...) on Linux machines/ GPU clusters, Git version control. Paper published on the MICCAI SASHIMI workshop (13 October 2019).