The challenge

Energy retailers and grid operators face a structural problem: the grid was built for a world of centralised, predictable generation that no longer exists. EVs, heat pumps, batteries and dynamic tariffs are multiplying faster than legacy systems can process, resulting in rising costs, operational fragility and a decarbonisation programme that is moving too slowly.

At the same time, energy retailers globally are under pressure to reduce cost-to-serve, launch low-carbon propositions faster and extract value from the device data their customers are generating – something legacy systems cannot do.

The solution

Rather than applying AI to one slice of the problem, Kaluza built a unified intelligence layer spanning customer data, device data, market data and billing. Built from the ground up as a data and intelligence engine, the platform embeds AI across billing, customer operations and energy optimisation – predicting demand, automating customer service, optimising flexible load and balancing the grid in seconds.

The platform is deployed by enterprise energy retailers and utilities including OVO Energy, ENGIE and AGL, and manages over 26 million contracted meters globally.

How the project works

Kaluza streamlines energy and decarbonisation in three ways. First, it automates billing, customer care and exception management. Second, it orchestrates smart charging, vehicle-to-grid (V2G), battery dispatch and heat pump optimisation across more than 30,000 devices, shifting load to clean, low-cost windows and reducing reliance on fossil-fuel peaking generation. Third, it enables retailers to launch new low-carbon propositions in as little as 19 days.

Recognising that AI systems carry their own energy and water costs, Kaluza has architected the platform to minimise unnecessary compute, running inference on pre-trained models rather than continuous retraining and co-locating intelligence with existing cloud infrastructure. Load is shifted to periods of clean energy availability wherever possible. The platform processes 5,000 API events per second with zero downtime, handling 24 million data points daily across its client base.

The results

Kaluza’s AI capabilities are delivering measurable, audited outcomes. On customer operations, AI-powered support tools have driven a 27% reduction in customer contacts and a 23% improvement in first-contact resolution, with Net Promoter Score improving by +39 across millions of customer accounts. Operational overhead has been reduced by up to 33%, and CRM and contact centre costs have been cut by £5m annually for individual clients.

On energy optimisation, the platform manages more than 200MW of flexible load and optimises 180GWh of flexible energy annually – equivalent to powering more than 20,000 homes for a year – while delivering more than £100 in value per customer per year through intelligent demand shifting. The world’s first and largest residential V2G programme, run with OVO Energy and Nissan, demonstrated 45% carbon savings compared with unmanaged charging, with customers earning up to £960 per year. Across its client base, the platform delivers £130m in annual technology cost savings.

Kaluza’s 2030 ambitions are 100 million contracted meters and 10GW of flexible load, which it says are achievable if more retailers decommission legacy systems and the market continues its shift away from monolithic CIS platforms. Future applications include whole-home energy management, trading optimisation and retailer-agnostic flexibility markets, enabling any device, on any tariff, with any retailer, to contribute to grid balance.