5+ years of experience in AI research or related fields
Strong background in machine learning and model architecture design
Proficiency in handling large datasets, particularly household-level data
Experience with rigorous evaluation methods and backtesting
Ability to communicate complex findings clearly in research papers
Responsibilities
Conduct pre-training research for the Energy Foundation Model
Design a unified model architecture incorporating diverse data inputs
Execute pre-training experiments to optimize model performance
Develop thorough evaluations and backtests with calibrated uncertainty
Ensure data privacy while training on sensitive household-level data
Collaborate with quants and engineers to transition models into production
Benefits
Competitive salary and eligibility for equity
Biannual bonus scheme
Fully expensed tech to match your needs
Private health insurance
Breakfast and dinner allowance for office-based employees
Full Job Description
As an AI Research Scientist, you will do the pre-training research behind the Energy Foundation Model: one model, trained on generation, the grid, demand, weather and price together, that learns the relationships between them that a human analyst would never think to write down. Today we model each forecasting problem separately; you will work out how a single model should be built and trained, and show that it works. We have data nobody else can buy, including household-level consumption from hundreds of thousands of UK homes, which outside trading firms cannot get. The work is research: experiments, evidence and papers that shape how we forecast, hedge and trade.
Requirements
Do the pre-training research for the Energy Foundation Model
Design the architecture, inputs and training objectives for one model across generation, grid, demand, weather and price data
Run pre-training experiments and find what actually improves results
Build rigorous evaluations and honest backtests, with calibrated uncertainty
Train on household-level data without leaking personal information
Write up findings as research, and work with our quants and engineers who take the model into production
Benefits
Competitive salary and eligibility for equity
Biannual bonus scheme
Fully expensed tech to match your needs
Private health insurance
Breakfast and dinner allowance for office-based employees
