Michal Uřičář
Senior AI Research Scientist at Rossum
Prague, Prague, Czechia
Summary
👤
Senior🎓
Top SchoolMichal Uřičář is a Senior AI Research Scientist based in Prague with 14 years of experience applying computer vision, machine learning and computer graphics to real-world products—from automotive perception at Valeo to speech research and business-document understanding at a specialist AI lab. He holds a PhD in Artificial Intelligence from Czech Technical University and brings hands-on systems expertise in Python, C/C++ and toolkits like PyTorch and TensorFlow, with notable low-level contributions such as implementing a Gauss–Seidel solver for Kernel Ridge Regression in the Shogun toolbox and core updates to an open-source facial landmark detector. Comfortable bridging research and production, he has international research experience including a Tokyo internship and a track record of shipping algorithmic improvements in C++ for performance-critical code. Colleagues rely on him for rigorous pattern-recognition solutions that balance theoretical depth with pragmatic engineering.
14 years of coding experience
5 years of employment as a software developer
PhD, Artificial intelligence and biocybernetics, PhD, Artificial intelligence and biocybernetics at Czech Technical University in Prague Faculty of Electrical Engineeringgeneral, general at Purkyňovo Gymnázium Strážnice
Italian, English, Czech, German, Russian
Stackoverflow
31reputation
7kreached
3answers
0questions
Github Skills (24)
algorithm10
algorithms10
c-language10
machine-learning10
data-structure10
data-structures10
cprogramming-language10
data-science9
lib9
code-library9
fileio9
file-access9
file-handling9
file-processing9
cpp8
Programming languages (3)
Github contributions (5)
uricamic/flandmark
Apr 2012 - Jun 2016
Open-source implementation of facial landmark detector
Role in this project:
Back-end Developer
Back-end DeveloperContributions:36 commits, 3 PRs, 1 push in 4 years 2 months
Contributions summary:Michal primarily focused on modifying and updating core C++ code within the flandmark repository, including a major version update. The commits indicate involvement in model loading/writing functionalities, and related file operations. Their work involved extensive modifications to `flandmark_detector.cpp`, indicating a focus on the underlying facial landmark detection logic, likely related to the model's core functionality.
deep-learningcomputer-visiondetectorlandmarkfacial
shogun-toolbox/shogun
Apr 2012 - Sep 2012
Shōgun
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:37 commits in 4 months
Contributions summary:Michal implemented a Gauss-Seidel iterative method for Kernel Ridge Regression within the Shōgun machine learning framework. Their work involved modifying the `KernelRidgeRegression.cpp` and `.h` files, introducing a new training method (`GS`) and related functions. Furthermore, the user updated the code to implement this new training method along with setting the epsilon value. They also fixed some comments and indentation in the code to improve readability.
cmakedata-sciencegunc-plus-plusmachine-learning
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.