Sr. Machine Learning Engineer
Summary
Senior ML Engineer developing and deploying deep learning models to automate orthodontic treatment planning, focusing on 3D geometric processing and multimodal data fusion using Python, PyTorch, and AWS.
- Decompose initial (poorly) stated business requests into separate technical problems and identify a specific ML toolset for a particular problem
- Apply machine learning approaches leading to improvements in orthodontic treatment planning
- Design and develop deep learning models for different data sources (e.g. 3D meshes, free-form text, table data etc.)
- Validate and analyse models performance
- Collaborate with the development team on architecture design of proposed solutions and support integration of the model into the production pipeline
- Document and transfer technology to product groups across the company
- Ownership of a particular ML component, its maintenance, and improvements
- Python
- PyTorch
- Docker
- Amazon Cloud
- Git/Bitbucket
- Master's degree or higher in Computer Science, Statistics, Machine Learning, Statistical Data Modelling or related fields
- 8+ years of experience in deep learning applications for commercial projects (preferably in computer vision tasks)
- Solid understanding of deep learning approaches, their limitations and powers (e.g. for semantic/instance segmentation tasks, Visual Transformers, modern LLMs, fusions of different data types, etc.)
- Proficiency in the standard technological stack: Python, git, Linux, Docker
- Mathematical background for solving ML tasks (optimization methods, Bayesian framework, linear algebra, analytic geometry)
- Strong experience in ML frameworks (preferably PyTorch)
- Experience in digging through the latest scientific papers and identifying applicability for a particular problem
- Strong interpersonal, oral, written, and visual communication skills, with ability to present findings concisely and effectively
- English B2 or higher
- Experience in 3D ML problems is a huge plus (e.g., pose estimation, scene reconstruction, surface/volume segmentation, etc.)
- Hands-on experience with in-depth exploration and customization of LLMs, including fine-tuning, embedding generation and optimization, retrieval-augmented generation (RAG) system development, and model evaluation, is a strong plus
- Scientific publications
- Experience in С++ is a plus
- Experience in ML Ops stack (e.g. Kubernetes, AWS cloud)
- Knowledge of SQL