What You'll Do
Design and deploy machine learning models on network data – including network selection optimization, churn and usage prediction, fraud and anomaly detection, and QoS/QoE forecasting.
Analyze CDRs, signaling events (Diameter, GTP, SIP), session logs, and radio-level metrics to identify performance issues, cost drivers, and growth opportunities.
Partner with the Network Operations and Product teams to translate raw telemetry into actionable signals – e.g., which carrier to steer traffic to in a given country at a given hour.
Build forecasting and segmentation models that inform pricing, capacity planning, and customer lifecycle strategy.
Own analyses end-to-end: framing the business question, exploring the data, building the model, validating it, and communicating findings to both technical and executive audiences.
Define and track KPIs for network quality, customer experience, and commercial performance, and build dashboards that make those metrics visible across the company.
Mentor junior data scientists and analysts, and help raise the bar on data science practices, code quality, and experimentation rigor.
What We're Looking For
5+ years of hands-on data science experience, ideally with exposure to telecom, networking, IoT, or large-scale event/log data.
Strong applied ML background: classification, regression, time-series forecasting, anomaly detection, clustering, and a working understanding of when to use what.
Expert SQL and strong Python (pandas, scikit-learn, PyTorch, or TensorFlow). Comfort with big-data tooling such as AWS Lakehouse, Redshift, and others.
Track record of shipping models into production and measuring their business impact, not just building notebooks.
Strong analytical storytelling – you can take a noisy dataset and turn it into a clear narrative for a non-technical stakeholder.
Telecom and networking concepts – cellular architecture (2G/3G/4G/5G), roaming, IMSI/IMEI, HLR/HSS, signaling protocols, QoS metrics. If you don't have this yet but have worked on similarly complex network/log data, we'd still like to talk.
Excellent written and spoken English.
Nice to Have
Familiarity with streaming data (Kafka, Flink) and real-time inference.
Background in pricing, revenue management, or unit economics analysis.
MSc or PhD in Computer Science, Statistics, EE, Physics, or a related quantitative field.
