ML инженер

Vidar Systems is a Ukrainian startup in the field of acoustic technologies, operating at the intersection of defense and security, developing acoustic systems that use sound to determine the location of hostile objects, including artillery and drones.

• Design, implement, evaluate, and optimize ML/DL models for acoustic detection and classification tasks;
• Collect, clean, and preprocess datasets for model training and evaluation;
• Write modular, testable code for preprocessing, model inference, and automation tasks;
• Develop and maintain robust data pipelines to support efficient model training and evaluation workflows;
• Develop strategies for model quantization and compression to ensure high performance on resource-constrained hardware.

• Bachelor’s or Master’s degree in Computer Science, Mathematics, Statistics, or a related field;
• 2+ Years of professional Python experience in ML/DL production applications;
• Hands-on expertise in designing, training, and integrating ML systems into production environments;
• Deep experience with TensorFlow, PyTorch, PyTorch Lightning, scikit-learn, NumPy, and Pandas;
• Strong analytical, communication, and problem-solving skills, with the ability to thrive and execute independently in a high-stakes, fast-paced startup environment;
• Good technical English;
• Nice to have: C/C++ experience, background in the audio domain or digital signal processing, proficiency in training unsupervised learning models, experience with the optimization and deployment of ML models on edge devices.

• High autonomy to shape the direction of software solutions with minimal oversight and maximum ownership;
• Innovation at the edge by contributing to military acoustic reconnaissance systems;
• Flexible work options with adaptable arrangements.