Summary
Daniel Coquelin is a Senior Software Engineer and Doctor of Engineering in Computer Science based in Karlsruhe with eight years of experience building and optimizing large-scale AI and high-performance computing workflows. He has driven algorithmic and systems advances—developing DASO for hierarchical data-parallel training, discovering a low-rank manifold stabilization used in OIALR, and creating AB training to slash communication by 70% while producing much smaller models—work accepted at major venues and benchmarked on Germany’s largest HPC systems. A core developer on distributed analytics frameworks like HeAT, he combines deep research rigor with hands-on engineering in MPI, NCCL, PyTorch and containerized deployments to deliver measurable speedups (including 20% network training gains and reported 1000x accelerations of legacy code). Comfortable consulting across academia and industry, he pairs hyperparameter optimization, energy-efficient container workflows, and practical performance tuning to move prototypes into production-ready pipelines. Colleagues rely on him for bridging cutting-edge ML research and production-grade distributed systems—an engineer who proves theoretical insights with compelling empirical wins.
8 years of coding experience
5 years of employment as a software developer
Doctor of Science, Artificial Intelligence, Doctor of Science, Artificial Intelligence at Karlsruhe Institute of Technology (KIT)
High School Degree, High School Degree at Annandale High School
Master of Science (M.S.), Physics, Master of Science (M.S.), Physics at The University of Bonn
Bachelor of Science (B.S.), Physics, Bachelor of Science (B.S.), Physics at James Madison University