Machine Learning Engineer

Data Science UA is a service company with strong data science and AI expertise. Our journey began in 2016 with uniting top AI talents and organizing the first Data Science tech conference in Kyiv. Over the past 9 years, we have diligently fostered one of the largest Data Science & AI communities in Europe.

About the role:
We are looking for an ML Engineer / Applied Scientist to join a multidisciplinary team working on a network intelligence and reliability project for a distributed wireless sensor network.
The role focuses on evaluating the feasibility and potential value of machine learning and advanced analytics for network-state understanding, degradation prediction, capacity forecasting, decision support and predictive maintenance.
You will work with real-world network telemetry, historical operational data, simulator-generated data and controlled field experiments.

Responsibilities:
ML & Data feasibility
- Assess the availability and quality of historical data for ML use cases
- Analyze data coverage, class/event distribution, temporal coverage and labeling quality
- Identify data gaps and determine whether additional instrumentation or data collection is required
- Evaluate whether the available data contains sufficient signal for predictive or classification tasks
- Define appropriate training, validation and test strategies for temporal and network data
- Avoid data leakage and ensure that evaluation reflects realistic deployment conditions
Network-state analytics
- Work with the Network Architect and Data Engineer to define an analysis-ready representation of network state
- Analyze network states based on variables such as: topology, node/link status, routing, RSSI/SNR, packet loss, latency, traffic/load, battery state, device availability
- Develop analytical and ML approaches for identifying healthy, degraded, constrained or unstable network states
- Investigate temporal patterns and transitions between network states
Baseline development
- Establish reproducible statistical and deterministic baselines
- Define appropriate evaluation metrics for each hypothesis
- Compare candidate ML models against simple baselines
- Quantify whether ML provides meaningful improvement over existing approaches
- Analyze trade-offs between accuracy, complexity, interpretability and implementation requirements
Simulator-based experiments
- Work with the Network/Embedded Engineer to use the existing network simulator for controlled experiments
- Analyze simulator-generated datasets
- Compare simulated and real-world data distributions
- Assess whether synthetic data can safely be used for model development and evaluation
- Support scenario generation and experiment design
- Evaluate whether simulator results generalize to observed real-world network behavior
- Help identify simulator calibration requirements

Requirements:
- 4+ years of professional experience in Machine Learning, Applied Data Science, Data Science Engineering or a related field
- Strong Python skills and hands-on experience with ML/data science workflows
- Strong understanding of: supervised and unsupervised learning; classification; regression; time-series analysis; anomaly detection; forecasting; model evaluation.
- Experience working with temporal and event-based datasets
- Strong understanding of data leakage, train/validation/test splitting and evaluation methodology
- Experience building reproducible ML experiments
- Strong statistical fundamentals
- Experience establishing and evaluating non-ML baselines
- Ability to work with imperfect, noisy and incomplete real-world data
- Ability to translate a technical/business hypothesis into a measurable ML experiment
- Ability to communicate model results, limitations and uncertainty to technical and non-technical stakeholders

Nice to have:
- Network / Telemetry / IoT
- Experience working with data generated by: wireless networks, IoT devices, sensor networks, telecommunications systems, distributed infrastructure, industrial systems, robotics, edge devices
- Experience with network telemetry such as: RSSI, SNR, packet loss, latency, link quality, routing, topology, network load, node availability, battery telemetry
- Graph & Network analytics
- Simulation & Synthetic data
- Forecasting & Reliability
- Experience with: degradation prediction, reliability modeling, predictive maintenance, capacity forecasting, anomaly detection, failure prediction, survival/lifecycle analysis
- Experience in telecommunications, IoT, industrial systems, robotics, defense or other physical-system domains would be a strong advantage.

Technical Skills:
Required
- Python
- NumPy
- Pandas
- scikit-learn
- SQL
- Git
- Jupyter / equivalent experimentation environment
Preferred
- PyTorch
- XGBoost / LightGBM
- Time-series libraries
- NetworkX
- MLflow or similar experiment tracking
- PyArrow / Parquet
- PostgreSQL / TimescaleDB or similar
- Cloud data/ML services