About Me
I am a Machine Learning engineer based in Switzerland. I’m interested in Computer Vision, Robotics, and building real-time software systems.
Projects
Product recognition in retail stores
Tech lead of product recognition team. See ShelfView Product website.
Barcode & text localization on mobile
Lead ML engineer for detection on edge devices. See MatrixScan Product website.
Virtual ads during live sports broadcasting
Image segmentation for real-time augmented reality rendering. See Viz Arena Product website.
Autonomous wheeled robot on cluttered grounds
Research on driving on obstacle avoidance for autonomous offroad driving. See The Gator autonomous vehicle website.
Experience
Senior Machine Learning Engineer | Scandit (Zürich, Switzerland) | 08.2021 - Present
- Tech Lead and Python Guild Lead for 40+ engineers
- Built large-scale object detection and classification pipelines, validated with US retailer testset
- Led visual search product development, scaled to >1M images/day with 10 customer pilots
- Increased end-to-end accuracy from 40% to 90% within 18 months
- Grew project squad from 2 to 7 engineers with mentoring and technical leadership
Machine Learning Engineer | Scandit (Zürich, Switzerland) | 10.2019 - 07.2021
- Launched real-time ML model for image segmentation running on 1M+ mobile devices monthly
- Trained binary neural networks for smartphones with 30x less memory usage
- Developed multi-stage training algorithms reducing model size by 50% at same quality
- Introduced Python coding guidelines and company-wide Request for Change process
Computer Vision Engineer | Vizrt (Zürich, Switzerland) | 10.2016 - 09.2019
- Developed real-time virtual advertisement overlay system for sports broadcasting
- Deployed neural networks for video segmentation in live TV with 50ms inference on 4K
- Automated training and fine-tuning of networks for customer-specific deployments
Robotics Research Intern | CSIRO (Brisbane, Australia) | 09.2013 - 08.2014
- Autonomous navigation research for large vehicles on industrial sites using stereo cameras
- Designed vision and LIDAR sensor fusion algorithms for obstacle avoidance