Paweł Kapica
Senior Machine Learning Engineer at Phoebe
Zurich, Zurich, Switzerland
Summary
👤
Senior🎓
Top SchoolPaweł Kapica is a Senior Machine Learning Engineer based in Zurich with over a decade of engineering experience and seven-plus years specializing in machine learning and AI infrastructure. He has co-founded an AI infrastructure startup, engineered a cost-saving platform (AXII) built from ~10 microservices and 235k lines of Go, and shipped production ML systems at Microsoft, Tractable, and startups. Paweł bridges research and production—moving models from prototyping to Kubernetes-powered deployment, building RAG pipelines for on-device code assistance, and creating ML observability tooling for computer vision. He contributes practical ML implementations (e.g., a TensorFlow YOLOv3 object detector) and thrives in cross-functional teams under tight timelines, pairing hands-on coding with product-driven impact. Known for optimizing operational costs and shipping end-to-end solutions, he combines deep model expertise (PyTorch/TensorFlow) with systems skills (Kubernetes, Go, distributed training).
11 years of coding experience
5 years of employment as a software developer
Master's degree Computer Science, Master's degree Computer Science at University of Warsaw
Polish, English, German
Stackoverflow
43reputation
2kreached
0answers
3questions
Github Skills (9)
object-detection10
deep-learning10
tensorflow10
python10
computer-vision9
machine-learning9
google-sheets6
padding6
java6
Programming languages (5)
Github contributions (5)
mystic123/tensorflow-yolo-v3
May 2018 - Jan 2022
Implementation of YOLO v3 object detector in Tensorflow (TF-Slim)
Role in this project:
ML Engineer
ML EngineerContributions:2 reviews, 47 commits, 12 PRs in 3 years 8 months
Contributions summary:Paweł primarily focused on implementing and refining the YOLO v3 object detection model within the TensorFlow framework. They contributed to the model's architecture, including the darknet53 backbone and the detection layers, and developed supporting functions such as upsampling, IOU calculations, and non-max suppression. The user also created a demo script to facilitate model usage and integration.
deep-learningobject-detectorobject-detectiondetectortensorflow
mystic123/io
Mar 2015 - Sep 2015
Contributions:60 commits, 52 pushes, 2 branches in 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.