About me

I am a passionate Deep Learning Engineer and Technical Leader, specializing in:

  • Machine Learning: Building and deploying state-of-the-art AI models.
  • Computer Vision: Driving advancements in real-time image processing.
  • Software Engineering: Designing scalable and efficient software systems.

Iโ€™m deeply interested in leveraging AI to help humans and accelerate scientific discoveries.


My Current Focus

  • ๐Ÿ“ Based in Lausanne, Switzerland.
  • ๐Ÿš€ Building BeEmotion.ai, a high-tech startup at the EPFL Innovation Park.
  • ๐Ÿ–ฅ๏ธ Exploring cutting-edge research combining LLMs with Computer Vision.
  • ๐ŸŒ Contributing to open-source projects to advance the tech community.

Technical skills

๐Ÿง‘โ€๐Ÿ’ป Tech Stack

  • Programming: Python, C/C++, Bash
  • Cloud & Orchestration: Google Cloud Platform, Kubernetes, Docker, Helm
  • ML Frameworks: PyTorch, TensorFlow, TFLite, TFX, ONNX, NCNN, TensorRT, Ray
  • Data Analysis: Numpy, Pandas, Scikit-learn, OpenCV, FiftyOne, CVAT, Neptune
  • Dev Tools: PyTest, Gitlab CI, Jenkins, FastAPI, MongoDB, ELK Stack, CMake

๐Ÿค– Deep Learning & Computer Vision

  • Multiple Object Detection & Tracking
  • 2D & 3D Human Pose Estimation
  • Headpose, Eye Gaze, Emotion Analysis
  • Unsupervised Learning for Time Series
  • Inference Optimization Techniques

Feel free to connect with me on LinkedIn or explore my projects on GitHub.


Short biography

Bartek Borzyszkowski is a Deep Learning Engineer experienced in commercial software development and academic research. He specializes in designing, implementing, and deploying efficient AI software across hardware platforms on the edge. Prior to joining BeEmotion.AI, he worked as a Research Engineer at EPFL and a Deep Learning Engineer at Intel AI. He completed his BSc and MSc in Robotics and Decision Systems at Gdaล„sk University of Technology, Poland. His experience includes also international research projects at CERN (Switzerland), USFQ (Ecuador), and the Karunya Institute of Technology (India).

EPFL Square