Case Study

Developing the right data science capabilities to support successful business transformation

Every day, water utility companies are tasked with providing drinking, waste and stormwater services to New Zealanders. Reliably supplying millions of litres of water and managing the challenges that come in hand with this is no mean feat. Having worked with a number of water utility companies, Harmonic understands these challenges and has the expertise required to help curate the right data science capabilities to address them.


In this case study, we take a high-level look at how a large New Zealand Water Utility company partnered with Harmonic to quickly set up and develop the right data science capabilities they needed to tackle a range of business challenges, laying the foundation for future growth.

The Challenge

As part of a major business transformation, a large New Zealand Water Utility company wanted to utilise analytics and insights to drive operational decision making and planning. To support this objective, they required increased data science capability to quickly develop and deploy proofs of concept projects coming from their change programme.

The Solution

Harmonic worked closely with the Utilities’ teams to provide data science team development. By helping with recruitment, setting a data science cloud platform and providing ongoing technical mentoring (data science best practices and expertise), Harmonic enabled the utility company to quickly build up to a team of five data scientists and accelerated solution development.

As data science capabilities have advanced, the team has tackled a wide variety of business problems together, including using machine learning to predict wastewater overflows and analysis of wastewater infrastructure performance in wet weather.

The Result

  • Gained data science resources required to provide data and analytics services to the wider business
  • Greater insights and improved forecasting capability to support operational and planning needs
  • Ability to quickly explore proofs of concept projects, and choose which to operationalise into business tools and dashboards
  • Gained confidence that internally produced work was accurate, in line with data science best practices and maintainable in-house.