Deep Learning Engineer

O projekcie: About Sybilla Technologies📡   Sybilla Technologies designs, builds and operates robotic optical observatories for monitoring objects in space. The company specializes in the complete systems, software for automated data acquisition, analysis and scheduling of networks of sensors. We work in the domains of Space Situational Awareness, Space Surveillance and Tracking, Space Traffic Management, stellar and planetary research, education, and commercial on-demand observations.    We are looking for a Deep Learning Engineer who is excited by the idea of training models that don’t just classify images but help observe and understand the sky itself. In this role, you’ll turn unique, real-world datasets — captured by a distributed, continuously operating sensor network — into advanced, intelligent systems. You’ll design and train deep learning architectures that push the boundaries of computer vision and time series analysis, delivering prototypes that shape the future of our products. If you enjoy building end-to-end ML solutions, experimenting with SOTA methods, and working closely at the frontier of science-driven engineering, this role will give you an extraordinary playground. Wymagania: Must haves: - 5+ years of proven experience in machine learning architecture design and solution delivery. - Excellent command of Python and deep learning frameworks (TensorFlow, PyTorch). - Strong expertise in computer vision (e.g., segmentation, object detection). - Experience with timeseries classification tasks. - Hands-on experience with image processing. - Practical experience in training and evaluating deep learning models. - Strong problem-solving skills and analytical thinking. Nice to Have: - Publications in AI/ML. - Experience training models on-prem infrastructure and defining its requirements. - STEM education background (optional). - Interest in Space-related topics Codzienne zadania: - Review current state-of-the-art deep learning algorithms in computer vision and time series classification. - Select and tune neural network architectures for specific tasks. - Analyze training datasets, including class distributions and statistics. - Prepare datasets for training (augmentation, filtering, and preprocessing). - Coordinate dataset generation with the Sensor Network Operations team. - Oversee training processes and hyperparameter tuning. - Continuously train and improve selected models. - Provide model explainability tools (e.g., saliency maps). - Develop MVPs and prototypes by demonstrating trained models. - Benchmark and evaluate ML models. - Support Data Analytics and Internal Tools teams with ML related analysis.