EPE is designing the grid of the future and needs to leverage AI and machine learning for load forecasting, distributed energy resources (DER) adoption and grid impact forecasting, and power system data analysis. This involves developing and implementing advanced modeling techniques to support electric and gas sales, rate making, regulatory functions, and operational analytics.
Requirements
- Expert in Python (pandas, scikit-learn, XGBoost, PyTorch/TF), PySpark/SQL, and cloud analytics (Azure Databricks or equivalents).
- Proficiency with hierarchical time-series, quantile regression, ensemble trees, Prophet, LSTM/Transformer networks, and scenario analysis.
- Familiarity with ISO/RTO market rules, IRP/DSIP filing requirements, DR/VPP measurement & verification, and grid operational constraints.
- Hands-on with PII minimization, Role-Based Access Control, encryption, and compliance standards in regulated environments.
- 8+ years in advanced statistics/ML on large-scale time-series; ≥3 years focused on utility load or DER forecasting.
- Experience with graph neural networks, physics-informed ML, and federated learning.
- Containerize models, implement CI/CD, and manage versioning (MLflow, SageMaker, Kubeflow).
Responsibilities
- Design, build, and maintain short-term (hours-to-days) and long-term (years-to-decades) load and DER forecasting models ranging from premise- to system levels of granularity.
- Integrate hierarchical, probabilistic, and physics-aware ML approaches; reconcile forecasts across network layers while honoring grid constraints.
- Model EV, PV, battery, DR, time-varying rates and heat-pump adoption curves and load flexibility; quantify coincident impacts and uncertainty bands for planning and real-time operations.
- Apply causal-impact methods (difference-in-difference, synthetic control) to measure demand-response and VPP performance.
- Ingest and curate sub-hourly AMI, SCADA, weather, and DER telemetry into cloud feature stores (Azure/AWS/GCP; Databricks, Delta/Parquet).
- Automate data-quality checks, lineage, and schema evolution for millions of devices.
- Containerize models, implement CI/CD, and manage versioning (MLflow, SageMaker, Kubeflow).
Other
- Translate methodologies, analysis, and results for utility planners, operators, and regulators.
- Present technical findings to executives, regulatory commissions, and cross-functional product teams.
- Mentor junior data scientists; publish or present at IEEE/NARUC forums; guide research sprints exploring graph neural networks, physics-informed ML, and federated learning.
- Proven ability to convert technical insights into clear business value and defend methods before both technical peers and non-technical stakeholders.
- BS/MS in Data Science, Statistics, Electrical Engineering, or related field; PhD, PE license, or peer-reviewed energy analytics publications preferred.
