Senior Machine Learning Engineer (Time Series) - Remote (across the UK only) - Climate Technology

Our client is driving a transformation in the commercial real estate sector by advancing sustainability through innovative retrofitting solutions. Focused on the idea that the most sustainable buildings are those that already exist, they provide commercial real estate owners with efficient, low-touch tools for optimizing their properties. By analyzing only building addresses, they identify tailored retrofitting strategies that reduce stranded asset risk, enhance property value, and secure favorable financing and partnerships.

As a Senior Machine Learning Engineer, you'll play a pivotal role in developing predictive models that assess and guide retrofitting efforts across various commercial buildings. You’ll design and implement machine learning solutions using energy, environmental and building data to forecast the impact of different retrofit measures, enabling clients to transition to sustainable building operations. This role offers opportunities for cross-functional collaboration with Data Scientists, Building Scientists, and other engineers, and allows you to contribute to projects such as natural language processing (NLP) for data workflows, cost modeling, and imagery analysis for property characteristics outside of the core projects that hone in on time series.

Key Responsibilities:

• Develop machine learning models to predict building performance improvements from retrofit actions, using detailed energy, environmental, and building data.
• Conduct feature engineering, especially with time series data, to enhance model accuracy and interpretability.
• Validate models rigorously to ensure performance and generalizability.
• Collaborate across disciplines to analyze building and energy behavior.
• Participate in code reviews and adhere to best practices for model deployment and production readiness.
• Contribute to broader data processes, including NLP for data ingestion, retrofit cost predictions, and imagery-based analysis of building features.
• Document and communicate technical insights and recommendations to both technical and non-technical stakeholders.

Qualifications and Experience:

Essential:

• Extensive experience in ML model development with deep expertise in applying machine learning to time series data, including:

1. Feature engineering for time series: Ability to extract meaningful features, handle missing data, and manage irregular time series
2. Data pre-processing for time series: Strong skills in cleaning, normalising, and transforming time series data, including handling seasonal trends and periodicities
3. Understanding of “classical” time series techniques, e.g. ARIMA, identification of seasonality, Kalman filters etc.
4. Deep learning for time series forecasting: Use of e.g. RNNs, LSTMs, GRUs, and Temporal Convolutional Networks (TCNs) for handling complex time series data.
5. Experience with multivariate time series: applying ML methods to multivariate time series data, capturing dependencies across different variables over time.

• Solid experience with Python and the associated libraries such as Pandas, Scikit-learn, TensorFlow, PyTorch, Darts, and sktime.

Bonus:

• Experience working with building energy data and a background in the built environment or sustainability.
• Familiarity with MLOps practices.
• Knowledge of geospatial and remote sensing data, and associated tools (e.g., Xarray, Dask, GeoPandas).
• A commitment to climate action and sustainable development.

If you're interested, please apply with your most recent CV and a consultant will be in contact. Please note that this role does not offer sponsorship at this time