Camille Driguez

◦ Supported a supply chain direction to scope and develop a multi-plant supply chain management application, currently used in production by 10+ users. ◦ Engaged with key users in workshops to design features and proposed a product roadmap ◦ Developed the python backend, a planning optimization algorithm to balance setup times with order delays at the plant level and a SAP integration FastAPI, Pydantic, Sqlalchemy

• Developed and deployed a service generating summaries from audio. Built a backend service orchestrating ML services. Prompt engineering ; developed a scoring pipeline to evaluate prompts and setup prompt A/B testing. Finetuned LLMs and worked on latency/cost optimization and inference acceleration. • Built and shifted into production a NLP API service for automatic medical data structuration from free text. Developed a first fully rule based API. Defined an annotation strategy and fine-tuned models.

• Implemented a NLP pipeline of text classification for activity risk scorings. • Geomarketing analysis to improve customer knowledge.

• Worked in a central team to identify, propose, scope, and deliver data science use cases with business impact for various teams across the group • Worked on a parameter setting recommendation model in production for product shift optimization of two glass coating lines that allowed the reduction of set up times up to 30%. • Built a churn model using a survival analysis bayesian approach • Leveraged dimensional reduction and clustering on quality data to identify defective product profiles revealing yearly savings • Developed sales forecasts and product targeting models using time series and dimensional reduction with an autoencoder for a distribution business unit