Machine Learning Engineer

What we do No matter if you’re the biggest sports fanatic, have a ‘minor’ shoe addiction or you own just one pair of sneakers, you know that finding the “perfect” shoe can be challenging. We’re here to help people find that perfect fitting shoe every single time. We have created a way to make a 3D scan of your feet and match that profile with our own (and the world’s largest) database of 3D-scanned shoes. We are also developing technology for 3D printing custom shoes and insoles based on a person’s foot profile. Our goal is to inspire people to do what they love in maximum comfort, whether it’s running a marathon, dancing all night or simply going for a stroll in the park. To accomplish this, we work with the largest shoe retailers and brands (Nike, HOKA, Intersport, Adidas, PUMA, Amazon and many more) across 70+ countries. As an employer, we strive to translate this philosophy into a happy and balanced lifestyle for all our team members, giving them the opportunity to grow and thrive while doing what they love. The fit we hope to find We're looking for an experienced Software Engineer with broad technical interests who wants to work across the full path from raw inputs and Computer Vision, through data processing and model training, to production inference and analytics. This is a software engineering role first: you should be comfortable designing, writing, testing and maintaining production-quality code, primarily in Python. Rather than owning one narrow component, you will help connect the pieces of the system - processing images, 3D scans and other data; maintaining data and training pipelines; improving ML/DNN models; deploying inference; and building analytics and diagnostics around the outputs. The scope is intentionally broad, and that is a core part of the opportunity. We are not looking for someone who is already an expert in every layer. If you are stronger in some areas and less experienced in others, don't let that put you off. What matters most is curiosity, learning speed, technical judgment and the willingness to dig into unfamiliar code, data and systems. You will have significant room to learn across ML, data engineering, Computer Vision and production software while developing deeper expertise over time What you'll do At SafeSize you will work across the end-to-end software chain behind our 3D scanning and recommendation technology - from camera inputs and Computer Vision processing, through data preparation and model training, to production inference, analytics and monitoring. You will not be expected to own every area alone, but you should be comfortable following a problem across boundaries and contributing wherever needed. As part of your role, you will: - Develop and maintain production-quality Python software across our ML, data processing, Computer Vision and 3D scanning systems; - Build and maintain data pipelines to ingest, clean, transform, validate and version the training, validation and test data used by our ML models; - Maintain and improve ML/DNN training pipelines, including reproducible experiments, training, validation, evaluation and the tooling that makes model development reliable; - Train, retrain or fine-tune models when needed; evaluate performance; perform error analysis; and help determine whether improvements should come from the model, the data or the surrounding software; - Package, version and deploy ML/CV models into production, and maintain reliable inference on-device or in the cloud; - Build data-processing and analytics workflows around scan data, model outputs and logs to understand model/product performance, diagnose problems and support better decisions; - Design and improve Computer Vision and image/3D processing pipelines for tasks such as feature extraction, camera-based measurements, calibration and geometric analysis; - Investigate real-world issues end to end using images, 3D scans, datasets, model outputs, logs and sensor data, and build diagnostics and tests that improve accuracy and robustness; - Build APIs, services and internal tools that connect data processing, ML, Computer Vision, inference and analytics with the rest of the SafeSize platform, working closely with Software, Data/ML, DevOps and Hardware teams; - Navigate unfamiliar systems and codebases when needed. You may occasionally need to read code in languages other than Python and, where useful, contribute to JavaScript/TypeScript-based tools or interfaces.