Pablo Ribalta
Director Deep Learning Algorithms at NVIDIA
Tychy, Silesian Voivodeship, Poland
Summary
🤩
Rockstar🎓
Top SchoolPablo Ribalta is a director-level deep learning engineer and researcher with eight years of industry experience driving high-performance model evaluation and large-scale training at NVIDIA. He builds and coaches teams that deliver production-grade deep learning across computer vision, medical imaging, drug discovery and graphics, while contributing to flagship open-source projects like NVIDIA DALI and DeepLearningExamples. His hands-on work spans performance optimization, mixed-precision inference, and GPU-accelerated data pipelines—practical skills underpinned by a PhD in AI and Machine Learning. Pablo’s background includes leading R&D for MRI-based diagnostics and hyperspectral imaging, producing peer-reviewed publications and patents, which gives him a rare blend of academic rigor and product-minded engineering. Based in Poland, he combines strategic program leadership (MLPerf/model evaluation) with day-to-day code contributions that noticeably improve training accuracy and throughput.
8 years of coding experience
13 years of employment as a software developer
Bachelor's degree in Computer Engineering, Computer Engineering, Bachelor's degree in Computer Engineering, Computer Engineering at Universidad de OviedoMaster's degree in Information Tecnology and Computer Engineering, Computer Engineering, Master's degree in Information Tecnology and Computer Engineering, Computer Engineering at Wrocław University of Science and TechnologyDoctor of Philosophy (Ph.D.), Artificial Intelligence and Machine Learning, Doctor of Philosophy (Ph.D.), Artificial Intelligence and Machine Learning at The Silesian University of Technology
Github Skills (21)
pytorch10
python10
machine-learning10
deeplearning-ai10
deep-learning10
tensorflow10
gpu10
data-processing10
computer-vision10
image-processing9
mlops9
dockers7
docker7
performance-tuning6
performance-analysis6
Programming languages (3)
Github contributions (5)
NVIDIA/DeepLearningExamples
Dec 2019 - Apr 2021
State-of-the-Art Deep Learning scripts organized by models - easy to train and deploy with reproducible accuracy and performance on enterprise-grade infrastructure.
Role in this project:
MLOps Engineer
MLOps EngineerContributions:19 commits, 16 PRs, 8 pushes in 1 year 4 months
Contributions summary:Pablo primarily focused on improving and maintaining the TensorFlow-based VNet segmentation model within the `nvidia/deeplearningexamples` repository. Their contributions include code cleanup, enhancing profiling methods, fixing typos, enabling mixed-precision inference, and removing a logging dependency. The user also updated checkpoint configurations and related scripts for MaskRCNN, demonstrating involvement in model training and deployment aspects.
forecastingcaffe2translationspeech-recognitionstate-of-the-art
NVIDIA/DALI
Sep 2018 - Nov 2018
A GPU-accelerated library containing highly optimized building blocks and an execution engine for data processing to accelerate deep learning training and inference applications.
Role in this project:ML Engineer
ML EngineerContributions:9 commits, 12 PRs, 5 pushes in 2 months
Contributions summary:Pablo's commits primarily involve modifications to the `nvidia/dali` repository, which focuses on data processing for deep learning. Their work includes fixing accuracy issues related to Tensorflow training, adding support for SSD in the COCO reader, and incorporating GPU versions of operators such as RandomBBoxCrop and Slice. Furthermore, the user has added and updated examples, specifically addressing thresholding in a Jupyter notebook detection example.
pythontrainingpaddlegpu-accelerationbuilding-blocks
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.