Summary
Marc Benzahra is a Senior Software Engineer specializing in AI/ML and NLP with 11 years of experience building production-ready models and data platforms across startups and consulting engagements. He combines deep academic research on surprisal-based readability and multilingual language modeling with hands-on MLOps, data engineering, and backend development to operationalize ML for business use cases. At DinMo he focuses on applying AI/ML to customer data within a modular CDP, and previously led LLM and productionization efforts for enterprise clients like Société Générale Factoring. His career spans research-driven roles (doctoral work and readability assessment at Glose/Medium) and pragmatic engineering (chatbots, entity recognition, and risk models), giving him a rare blend of theoretical insight and production discipline. Fluent in English and French contexts, he has repeatedly moved models from prototype to private-cloud deployment and automated pipelines. An interesting thread through his work is using surprisal and unsupervised signals to create practical, language-agnostic solutions that reduce reliance on labeled data.
11 years of coding experience
2 years of employment as a software developer
Epitech
Doctoral Student Computer Science (Natural Language Processing, Machine Learning), Doctoral Student Computer Science (Natural Language Processing, Machine Learning) at Université Paris-Saclay
Master of Science (MSc) Computer Science and Artificial Intelligence, Master of Science (MSc) Computer Science and Artificial Intelligence at University of Kent
French, English