Yves Jean — Co-Founder

Senior software engineer and computer vision research scientist who enjoys finding novel solutions to computational problems, writing software and building custom distributed systems. As a founder of kodou.io and CTO, I have built production-grade developer tools, from multi-language compiler infrastructure through runtime orchestration, and hands-on debugged distributed systems. Recognized expertise from distributed systems, and enterprise applications to CUDA, Bayesian inference (MCMC), and computer vision (https://github.com/yvesnyc). Experience includes contributions to developer tools (compilers, runtimes), context-controlled local LLM systems, and hands-on experience debugging and scaling distributed systems. Extended open source tooling such as Clang/LLVM, Tycho compilers, and Akka. The resulting kodou.io platform is industry-recognized (https://akka.io/customer-stories/kodou-solves-enterprise-technical-debt-with-innovative-akka), demonstrating my ability to architect systems that scale. Published articles in computer vision, camera models, and spatial coded light research in CVPR, ECCV, ICPR, Journal of Computer Vision and Image Understanding. Full list of publications on Google Scholar https://scholar.google.com/citations?user=mXAWTJQAAAAJ My work in super-resolution scene understanding includes CUDA-implemented algorithms built upon convolutional operators, coupled with coded light and illumination embedded filters in a projector-camera system (Github/yvesnyc). This work unifies scene geometry and camera models within a single computational framework. I experiment with the SAM2 vision model, but I believe the real potential of these systems lies beyond points and bounding boxes, as demonstrated by recent work such as VGGT (Visual Geometry Grounded Transformer), and my research on scene geometry computation and understanding aligns with this direction. This research-to-production capability stems from deep systems experience. These complementary skills, published research and software development, are best exemplified by my project at Bell Laboratories, as part of a two-person team, transformed our computer vision laboratory research into a real-time broadcast technology for professional sports. Today, this computer vision application is the standard for professional tennis broadcasts. Specialties: Computer Vision, Vision models, Novel Cameras, Java, C, C++, Scala, SAM 2, PyTorch, CUDA

Stackforce AI infers this person is a SaaS architect with expertise in distributed systems and computer vision.

Location: New York, New York, United States

Experience: 35 yrs 11 mos