Xavier Dupré
Teacher In Computer Science at Microsoft
Paris, Ile-de-France
Summary
🤩
Rockstar🎓
Top SchoolXavier Dupré is a versatile computer scientist and educator with a PhD and over two decades of experience combining academic teaching at ENSAE/ENSAE ParisTech and industry roles at Microsoft, Yahoo, Autonomy and startups. He blends deep ML and search expertise—having worked on Bing relevance, Azure ML, and query rewriting—with hands-on engineering in Python, C++ and large-scale data platforms. An experienced instructor, he authored widely used Python teaching materials and runs courses on machine learning, distributed algorithms and actuarial Python. His open-source contributions include improving TensorFlow-to-ONNX conversions for image models, adding quantization and complex operator support, reflecting practical ML deployment know-how. Based in Paris, he uniquely bridges research, production ML engineering and pedagogy, mentoring students through project-based evaluation and real-world tooling.
12 years of coding experience
14 years of employment as a software developer
Sup/Spé M', Mathematics, Sup/Spé M', Mathematics at Lycée Saint-LouisDoctor of Philosophy (Ph.D.), Computer Science, Doctor of Philosophy (Ph.D.), Computer Science at Université Paris VMaster, Statistics, Finance, Computer Science, Economy, Insurance, Master, Statistics, Finance, Computer Science, Economy, Insurance at Ecole nationale de la Statistique et de l'Administration économiqueMaster, Machine Learning, Artificial Intelligence, Master, Machine Learning, Artificial Intelligence at Université Paris VI
French, English
Github Skills (6)
quantization10
model-conversion10
tensorflow10
onnx10
machine-learning9
image-processing8
Programming languages (7)
Github contributions (5)
onnx/tensorflow-onnx
Jul 2020 - Aug 2021
Convert TensorFlow, Keras, Tensorflow.js and Tflite models to ONNX
Role in this project:
ML Engineer
ML EngineerContributions:181 commits, 1 comment in 1 year 1 month
Contributions summary:Xavier primarily contributed to the conversion of TensorFlow models, specifically focusing on models designed for image processing tasks. Their commits involved replacing the operator FakeQuantWithMinMaxVars with a QuantizeLinear and DequantizeLinear operator. Further contributions included creating unit tests for the quantization process, and adding support for operators like atan2, and improving the handling of the reshape operation, and the handling of operators such as RFFT and ComplexAbs.
tensorflowjsexportdeep-learningonnxkeras-tensorflow
sdpython/botadi
Dec 2018 - Jun 2022
Contributions:75 pushes, 1 branch in 3 years 6 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.