The UK’s energy system is under unprecedented strain: electrification of heat and transport, volatile wholesale prices, and surging peak demand create pressures that traditional tariffs are unable to resolve. EDF’s mission is to help Britain reach Net Zero, but doing so requires more than generating clean power. It demands smarter consumption, with households actively shifting demand in ways that support both grid stability and carbon reduction.
EDF needed a way to empower customers to save money while also reducing pressure on the grid. The answer lay in data: leveraging smart meter readings and AI models to create products that reward behavioural change in real time.
“This is a great example of change for measurable and significant outcomes, building on existing capability to evolve.” – Judges’ comments
Flagship Flexibility
The EDF UK Data and AI team developed two flagship demand-side flexibility programmes:
- Sunday Saver: Customers who reduce weekday peak-time consumption (typically 4-7pm) can earn up to 16 hours of free electricity on Sundays. By analysing smart meter data, EDF measures usage shifts and calculates tailored rewards, offering a clear, tangible benefit for changing habits.
- Beat the Peak: Linked to the National Grid’s Demand Flexibility Service, this initiative invites households to cut usage during defined peak events. AI models predict and assess consumption patterns, and participants are credited based on reductions achieved. Now in its second year, the scheme has seen more than 139,000 households participate.
Both products are underpinned by a robust data architecture. Snowflake’s Data Cloud provides a centralised, secure repository for customer and smart meter data. Amazon SageMaker, integrated with Snowflake, enables rapid development and deployment of AI models, reducing time-to-production from months to weeks. With half-hourly smart meter readings feeding the system, EDF can run near real-time analysis and adjust programmes dynamically.
Tangible Outcomes
The impact has been significant across three dimensions:
- Customer Engagement and Behaviour Change: Sunday Saver participants have collectively received 6 million hours of free electricity, worth £1.4m. Customers shifted as much as 40% of their peak-time usage, and EDF’s campaigns generated 31.2m social interactions, demonstrating both reach and resonance. Beat the Peak participants earned an average of £12.56 across nine events, with top households earning £51.80.
- Environmental Benefits: Collectively, Beat the Peak saved 23 tonnes of CO₂, the equivalent of driving 57,500 miles in a petrol-fuelled car. These reductions, while modest individually, highlight the aggregate potential of AI-enabled demand flexibility at national scale.
- Operational Efficiency and Scalability: By integrating Snowflake and AWS, EDF has streamlined its data operations. The infrastructure allows the team to scale initiatives quickly, adapt to shifting market conditions, and personalise customer engagement.
A Model for Energy Transformation
The judges noted EDF’s innovation in its ability to turn AI into simple, compelling propositions for households. These initiatives prove that AI can be a bridge between corporate sustainability goals and customer incentives, aligning cost savings with carbon savings.
For EDF, winning Transformation with AI (Europe) reflects a broader ambition: to make data and AI an engine of Net Zero. With dozens of AI products now in play, Sunday Saver and Beat the Peak stand out as transformative proof points, showing how AI can rewire behaviour at scale for a cleaner, smarter energy future.
