Industrialising ML use cases with a central Data & Analytics platform (“NEO”) and long-lived business teams. Datafy is at the heart of NEO.

Business context

Luminus is the second largest electricity producer and energy supplier on the Belgian energy market.

In 2019, Luminus successfully started experimenting with machine learning, but had no experience with operating machine learning use cases. The Luminus’ technology stack was built for traditional business intelligence, on on-premise infrastructure.

Scope & objectives

Set-up an operating model and cloud-native data & analytics platform which allows to build and run machine learning applications, leveraging on self-service.

Key results

Implementation of a new IT-oriented Landing Zone and Data & Analytics Platform capabilities​, based on Data Minded Cloud. It covers the full machine learning workflow from experimentation to monitoring, and required design patterns (batch processing; streaming).

We have set up long-lived business-oriented use cases teams, including new roles across business lines and IT. This new way of working helped to establish a culture of increased ownership, innovation and learning.

Impact

Within one year, 25+ users (scientists and engineers) were on-boarded, 7 use cases went live; 18 projects have started; 21 data sources were connected.

Positive ROI after delivery of first use cases.

Widespread adoption of data platform in different business lines.

Continued value-driven industrialisation of new use cases.

Strong cultivation through Data Minded Academy.