Kaluza provides an intelligent platform for the future of energy by revolutionising how the world’s biggest energy utilities serve customers. As energy markets worldwide become more decentralised and renewable energy generation ramps up, Kaluza’s utility clients are increasingly reliant on data to understand their broad spectrum of customers and provide personalised, innovative propositions and customer support. 

Kaluza’s real-time cloud platform is transforming retail operations, providing utilities with advanced solutions for automating their core operations, optimising energy flows and bolstering customer lifetime value through rewarding, low-carbon propositions. 

As a way to further enhance the value provided to their clients, Kaluza has partnered with Indicium AI’s team of experts to leverage AI & machine learning (ML) technologies. The partners engagement has focused on utilising ML models to predict an energy customer’s propensity to churn, with a view to develop new techniques to increase customer loyalty and lay the foundation for clients to maintain revenue growth.

The Challenge: Predicting Customer Churn To Maximise Customer Stickiness 

Kaluza wanted to develop a model that could be utilised to predict a customer’s propensity to churn, either as a result of a home move, or a customer changing supplier. This use case was chosen for a number of reasons:

• After a turbulent time within the global energy market, customer lifetime value and the ability to effectively manage customer churn is a priority for energy retailers. 
• Customer insight is crucial in order for contact centre agents to drive proactive next best action and the likelihood that a customer might leave is a crucial signal to drive retention activity.
• Building on five years of energy software innovation, Kaluza wanted to explore the opportunity to partner with Indicium AI, known for being on the cutting edge of AI innovation, to develop additional AI & ML capabilities as part of their energy retail proposition. 

The Solution: Harnessing Expertise for an Integrated ML Model

Indicium AI, in partnership with AWS, successfully delivered a 4-week project aimed at showcasing enhanced customer insights integrated into Kaluza’s core platform. This helps to drive next-best-action for clients, and acts as a starting point for future AI innovation. Leveraging Indicium AI’s data analysis, engineering and machine learning expertise, the team delivered a model that can be utilised by clients worldwide. Over the course of four weeks, the team was able to:

• Perform exploratory data analysis on key datasets to first examine data completeness and quality.
• Introduce an iterative process to develop and improve the model, utilising industry-leading techniques and technologies.
• Develop a clear methodology and narrative to support churn insights so that Kaluza’s utility clients could easily understand. 
• Build a pathway to move future use-cases forwards from ideation to demonstrable capabilities.

Deploying AWS Technologies

As an Advanced Partner, Indicium AI has a well established AWS practice, with capabilities across the AWS lifecycle. Indicium AI and Kaluza deployed a range of AWS technologies throughout the engagement which could be utilised in the future to drive long-term value for Kaluza customers:‍

• Data was split into raw form and processed for storage on AWS S3 Bucket, based on its use for querying or ML modelling.
• The team made data available for data science workspaces and compute via SageMaker, allowing for exploratory data assessments, prototyping, training and deployment workflows.
• Utilised CodePipeline for continuous integration and deployment, and CodeCommit as the code repository. 

The Value to the Business

The successful delivery of the model clearly demonstrates Kaluza’s ability to deliver AI & ML use cases in an accelerated time frame, to a high level of quality and accuracy. The engagement has led to the sharing of ideas and expertise between the Kaluza data team and Indicium AI’s team, further strengthening Kaluza’s flexible and modular platform. 

This demonstrable ML model now allows for the rapid deployment of churn insight for clients, via an API endpoint. This integration also aims to promote dialogue with Kaluza’s energy retail clients about future feature enhancements like agent recommendations, further next best action and ultimately, pave the way for revenue growth by maximising customer lifetime value.