Time-Series Forecasting Engineer (Load / Demand)

Job Title: Time-Series Forecasting Engineer Load / Demand (1 Position)

About the Job
We are looking for a pragmatic, data-driven Time-Series Forecasting Engineer to join our analytics team and lead the design, implementation and productionisation of short- and medium-term load/demand forecasting systems. You will build robust, explainable and probabilistic forecasting pipelines that drive operational planning, predictive maintenance, energy scheduling and commercial decisions across our UPS, battery and power systems. This role sits at the intersection of data science, software engineering and MLOps you will own models from data ingestion to deployment and monitoring in production.

Responsibilities
- Lead end-to-end development of time-series forecasting models (deterministic & probabilistic) for load, demand and related signals.
- Design data ingestion and feature engineering pipelines for telemetry, SCADA/PLC logs, environmental sensors and external covariates (weather, calendar, events).
- Implement statistical and ML approaches: ARIMA/ETS, state-space models, exponential smoothing, Prophet, classical ML, gradient boosting (XGBoost/LightGBM), RNNs/LSTMs, and modern Transformer/Temporal Fusion/DeepAR architectures.
- Produce probabilistic forecasts (prediction intervals, quantiles), calibration, and uncertainty quantification for risk-aware decisions.
- Build reproducible training, validation and backtesting workflows with robust evaluation metrics (MAE/MAPE/SMAPE/RMSE, CRPS, Pinball loss, sharpness/calibration).
- Optimize models for latency and throughput; prepare lightweight versions for edge/cloud inference (ONNX, quantization) when required.
- Deploy and monitor forecasting services (CI/CD for models, drift detection, automated retraining, alerting, SLA monitoring).
- Collaborate with product, operations and grid/planning teams to translate forecasts into actionable schedules and KPIs.
- Document models, assumptions and performance; present results to non-technical stakeholders and recommend operational changes.

Qualifications
- BSc/MSc in Data Science, Statistics, Computer Science, Engineering, Applied Math or related field (PhD appreciated but not required).
- 3+ years professional experience building and deploying time-series/forecasting models in production (energy, grid, utilities, manufacturing, or related domains preferred).
- Deep practical experience with time-series methods, probabilistic forecasting and sequence models.
- Strong Python skills and familiarity with libraries such as pandas, scikit-learn, statsmodels, PyTorch/TensorFlow, Prophet, GluonTS or equivalent.
- Experience with MLOps tooling (Docker, Kubernetes, MLflow, DVC, Airflow) and cloud platforms (AWS/GCP/Azure).
- SQL proficiency and experience working with large telemetry datasets.
- Solid software engineering practices: unit tests, CI/CD, reproducible experiments and versioning.
- Strong analytical thinking, experimental rigor and ability to explain tradeoffs to stakeholders.
- Excellent communication skills and ability to work cross-functionally.
- Bonus: domain experience with UPS/battery systems, energy markets, demand response, or forecasting for distributed energy resources.

What We Offer
- Ownership of a high-impact forecasting product used for operational and commercial decisions.
- Collaborative, cross-disciplinary team (data engineers, ML engineers, product and operations).
- Access to large real-world telemetry datasets, lab equipment and cloud compute resources.
- Competitive salary, flexible work arrangements and professional development support.
- Opportunity to shape our predictive analytics roadmap and scale industry-grade solutions.

How to Apply
Please submit the following to [Confidential Information] with subject line Time-Series Forecasting Engineer Load / Demand:
1. CV (max 2 pages)
2. Cover letter (1 page) summarizing relevant experience and interest in load/demand forecasting
3. One brief project summary (200300 words) describing a forecasting project you led: data, model(s), evaluation metrics and outcome
4. Links to portfolio / GitHub / notebooks (if available)
5. Two professional references (name, role, contact)

Shortlisted candidates will be invited for a technical interview and may be asked to present a short case study or practical task. We welcome applicants from diverse backgrounds and encourage practical, production-focused experience.