We are considering candidates who are located in Ukraine!
We are looking for an experienced AI/ML Model Specialist to join a product engineering team. In this role, you will work as part of the core engineering group, focusing on developing, optimizing, and deploying machine learning models for real‑world applications. The position involves close collaboration with engineers, data scientists, and domain experts to build reliable and scalable ML solutions using both classical machine learning and modern deep learning methods. Minimum Qualifications: * Master’s or PhD in Computer Science, Machine Learning, Engineering, Applied Math, Physics, or a related technical field * 3–5+ years of experience building and deploying ML models in production environments * Strong expertise in classical ML algorithms such as XGBoost, Random Forest, SVM, k‑NN * Strong understanding of deep learning architectures including CNNs and Transformers * Proficiency in Python along with libraries such as scikit‑learn, NumPy, pandas * Familiarity with C++ * Hands-on experience with multi-class classification using real-world and noisy datasets * Strong understanding of statistics, model evaluation, and validation methods * Upper-Intermediate level of English.
Preferred Qualifications: * Experience working with sensor or time‑series data including magnetic, radar, 3D, or IoT sources * Experience with feature extraction techniques such as FFT, windowing, and frequency‑domain analysis * Background in handling imbalanced datasets and label quality issues * Knowledge of feature interpretability methods and tools * Experience with MLOps tools such as MLflow, Weights and Biases, and CI/CD pipelines for ML * Experience monitoring and managing model drift * Experience optimizing ML models for edge devices.
Key Responsibilities: Feature Engineering and Sensor‑Aware Modeling — Design, develop, and optimize ML models using algorithms such as XGBoost, Random Forests, SVMs, CNNs, and Transformers — Perform hyperparameter tuning, feature selection, and algorithm evaluation — Build reproducible training pipelines with proper model, data, and experiment versioning — Extract temporal, spectral, and domain-specific features from raw sensor inputs — Apply analytical techniques such as UMAP and T‑SNE for data exploration — Model sensor-specific behaviors including noise, bias, drift, and environmental effects — Conduct ablation studies and feature importance analysis using tools such as SHAP and PDP.
Multi‑Class Detection and Classification — Design ML pipelines for multi‑class object detection and multi‑class classification in noisy and imbalanced environments — Develop evaluation metrics such as confusion matrices, calibration metrics, and class‑wise performance scoring. What you will love about working here? * We care about all our employees and want them to feel as comfortable as possible. That’s why we offer them health insurance from the first days, regardless of the probationary period. * The gift from the company — Christmas holidays from 25 December to 31 December. * Сooperation with Superhumans center and Veteran HUB. Capgemini Engineering has supported the launch of psychological rehabilitation department of Superhumans. Our team also donnated over UAH 500 000 prosthetics for three Ukrainian defenders. Currently, we support psychological counseling provided by the Veteran Hub, and we have implemented a internal policy making the company friendly to military and veterans with the assistance of the Hub.
