ML Engineer — Energy Grid Optimisation

Finn Larsen

Contract-to-hire · Mid-level · London

This role is no longer accepting applications

Similar open roles you can apply to now:

About the role

We are an electricity distribution network operator. We manage the infrastructure that delivers power to 3.4 million homes and businesses across a significant portion of the UK. We are not a technology company. We are an infrastructure company — one that has been operating for 40 years and intends to operate for 40 more. We say this at the start because we think it is important context for the kind of organisation you would be joining, and we have found that candidates who thrive here are ones who understood what they were walking into rather than arriving with different expectations. What we are doing, slowly and carefully, is applying machine learning to the operational problems that have resisted conventional approaches: load forecasting at substation level, fault prediction from network sensor data, and optimising the scheduling of maintenance activities to minimise outage risk and customer impact. We have a small analytics team that has been doing good work with conventional statistical methods for several years. We are ready to bring in ML engineering expertise on a contract basis, with the intention — subject to performance and budget — of converting to a permanent role. The work requires someone who is comfortable doing technically solid engineering in a domain where the data is slow-moving, the consequences of error are real, and patience with organisational process is a genuine job requirement.

Responsibilities

Build and validate load forecasting models at substation granularity using historical demand data, weather feeds, and network topology
Develop fault prediction models from network sensor telemetry and historical fault records
Contribute to the optimisation of maintenance scheduling by providing ML-based risk signals to our asset management team
Integrate ML outputs into the team\'s existing analytical infrastructure — this is not a greenfield project, it is an extension of working systems
Document all models, data pipelines, and evaluation methodology in a form that can be reviewed by our internal engineering governance process

Requirements

4+ years of ML engineering with experience in time-series forecasting, anomaly detection, or predictive maintenance in an industrial or infrastructure context
Python at a professional standard — clean, tested, documented code that an operations engineer can read and a data scientist can extend
Scikit-learn for time-series modelling, feature engineering, and the classical statistical approaches that remain the right tool for many of our problems
Apache Spark for processing the network telemetry and SCADA data volumes that our grid produces
SQL for working with our asset management database, outage records, and customer impact data
Statistics at a practical engineering level: regression, time-series decomposition, uncertainty quantification, and the humility to choose a linear model over a neural network when it is the right call
Familiarity with operational technology environments — SCADA, ICS, or industrial IoT — is an advantage but not required

Benefits

£65,000 – £82,000 annualised contract rate — permanent salary will be discussed at conversion
6-month initial contract with strong intention to convert to permanent
Hybrid — two days per week in our regional office, three days remote
Pension contribution of 7% employer match at permanent conversion
Job stability that is unusual in technology roles — we have not made compulsory redundancies in 15 years
Work on infrastructure that is genuinely critical — the problems are real and the stakes are clear