Summary
Mohamed Chelali is a postdoctoral researcher and applied AI scientist with 11 years of experience specializing in computer vision, image processing and multimodal AI for large-scale photographic archives. He blends academic rigor—doctoral work on spatio-temporal sequence analysis with multiple international publications—with product-focused engineering, having built video summarization engines and internal datasets for platform optimization at Jellysmack. At CNRS he develops multimodal pipelines to extract metadata, transcribe embedded text and structure cultural heritage corpora for humanities research, demonstrating a rare fluency between SHS needs and cutting-edge vision models. He is pragmatic about data quality—“no data == no science”—and leverages annotation tools like Arkindex to turn messy historical collections into research-ready datasets. Based in Paris, he teaches and mentors across programming and image analysis, and continues to pursue technology scouting to anticipate new AI opportunities. An understated strength is his ability to translate scholarly problems into deployable ML prototypes that serve both researchers and product teams.
11 years of coding experience
2 years of employment as a software developer
MLOps, Computer Programming, MLOps, Computer Programming at DataScientest.com
Doctor of Philosophy - PhD, Informatique, Analyse de séquence d'images, Doctor of Philosophy - PhD, Informatique, Analyse de séquence d'images at Université Paris Descartes