Senior Data Scientist – Payload & SAT-RAN

AST SpaceMobile•Lanham, MD

7d•Onsite

About The Position

AST SpaceMobile is building the first and only global cellular broadband network in space to operate directly with standard, unmodified mobile devices based on our extensive IP and patent portfolio and designed for both commercial and government applications. Our engineers and space scientists are on a mission to eliminate the connectivity gaps faced by today’s five billion mobile subscribers and finally bring broadband to the billions who remain unconnected. The Senior Data Scientist – Payload & SAT-RAN defines, builds, and operationalizes SAT-RAN data models across Payload, Gateway (GW), and Cellular/RAN subsystems. This role focuses on creating robust telemetry-driven models and metrics that quantify system performance, detect and prevent failures and anomalies, and enable data-driven optimization of user experience, capacity, and duty cycle under various ground and in-orbit constraints. This person works closely with payload engineering, gateway/network engineering, RAN/system architects, and operations to translate complex, multi-domain data into actionable insights and production-grade analytics.

Requirements

Bachelor's or Master's degree in Computer Science, Electrical Engineering, Statistics, Mathematics, Physics, or a related technical field, or equivalent experience.
5+ years of experience delivering data science/ML solutions in production, including monitoring, anomaly detection, forecasting, quality scoring, or optimization support.
Strong proficiency in Python (pandas, NumPy, scikit-learn, and time-series tooling) and strong SQL skills.
Demonstrated experience defining and operating data models and analytics pipelines, including schemas, identifiers, aggregation logic, data validation, and lineage.
Strong statistical foundations, including model evaluation, uncertainty, time-series behavior, bias/variance, and backtesting.
Ability to translate cross-domain system problems into measurable metrics and deployable analytics.

Nice To Haves

Experience with multivariate anomaly detection at scale, including change-point detection, sequence models where justified, and graph/topology-aware features.
Telecom/systems experience, including LTE/5G KPIs/KQIs, OSS counters/alarms, QoE/QoS metrics, and RAN performance indicators.
Familiarity with scheduling/capacity modeling concepts, including resource allocation, interference-aware capacity, constraint modeling, and power/thermal-limited regimes.
MLOps experience, including deployment, monitoring, drift detection, and CI/CD for data and models.

Responsibilities

Own Payload/GW/SAT-RAN data strategy: define key subsystem metrics, data sources, and collection requirements spanning payload, gateway, transport, and RAN/service layers.
Design and maintain SAT-RAN subsystem data models (entities, relationships, identifiers, etc.) to unify telemetry across domains and support scalable analytics.
Lead data model definition and deployment into production systems, including instrumentation requirements, pipeline design, validation, versioning, and data quality monitoring.
Develop performance analytics for end-to-end Satellite–RAN performance projected across multiple subsystems, including attribution of impact across payload, GW, transport, core, and RAN layers.
Build failure/anomaly detection and prevention systems using multivariate time-series, correlation/causality-informed approaches, topology/context-aware features, and alert deduplication/triage scoring.
Create qualitative quality scoring models that combine predictive signals with measured KPIs/KQIs, including confidence/uncertainty measures.
Develop fleet service scheduling models linking orbital state/visibility, predicted SAT-RAN capacity and quality, interference, and spacecraft power/thermal constraints to achievable service performance and demand fulfillment.
Build decision-support data products: dashboards, health scores, early-warning indicators, incident enrichment, and executive-ready reporting for SAT-RAN performance.
Partner with engineering teams to define success metrics, run backtesting/regression evaluation, and operationalize models into workflows such as assurance, release validation, and optimization loops.
Establish best practices for reproducibility and MLOps, including model monitoring, drift detection, dataset/version governance, and documentation.