EnergyAustralia is one of Australia’s largest energy retailers, with a business that spans retail energy customers, generation assets, and wholesale trading. The company’s purpose is to lead and accelerate the clean energy transformation for all, and its leadership sees data and AI as critical to making that transition work, both commercially and operationally.
But Australia’s wholesale energy market moves fast: it settles every five minutes, prices can be highly volatile, and new battery and sensor data arrives at far higher frequency than traditional systems were built to handle. In this environment, the ability to move quickly from prototype to production isn’t just a technical advantage – it's a commercial imperative. Pricing models and forecasting methodologies need to evolve continuously to reflect changing market dynamics, and the cost of slow delivery is real. Outdated models can lead to pricing inaccuracies, poorer forecast accuracy, and increased financial exposure during periods of volatility.
To keep up, EnergyAustralia continues to modernise its data foundation, starting with the Trading domain. Databricks is one of the platforms supporting this work, giving teams somewhere to build, test, and deploy new models.
Fast, reliable data is the is the foundation for commercial decisions
Australia’s wholesale energy market is highly volatile. Wholesale electricity spot prices can range from negative values to over $15,000/MWh in each individual five-minute period, depending on renewable output and demand conditions. In this environment, the speed at which teams can access and analyse data directly impacts the quality of commercial decisions – delayed insight is expensive.
Michelle Shek, Wholesale Pricing Leader, put it simply: “Having reliable data at our fingertips quickly is critical. It helps us to analyse the market and make sound commercial and financial decisions.”
Databricks helps to provide EnergyAustralia with a solid foundation to bring together five-minute settlement data, battery telemetry, and generation inputs in one place. This gives teams a trusted base to run pricing, hedging, and forecasting activities. Rather than waiting several hours for data to become available and analytical jobs to run, teams can now access what they need in close to real time.
The Databricks features available to EnergyAustralia are cutting edge. They enable users to grow their capabilities and leverage more out of what the data has to offer. Databricks apps, Genie, and Unity Catalog are just a few features that are enabling a better data future at EnergyAustralia.
Empowering teams to build and develop applications, not just consume
For EnergyAustralia, self-service means more than access to dashboards. It means giving business teams ownership of the full lifecycle, from developing applications and models through to deploying them in production.
As Chirag Shah, Trading, Data and Systems Leader put it: “What Databricks has enabled us to do is to do things ourselves.”
The wholesale pricing team, which is primarily business users rather than developers, now writes SQL and Python directly. They build and adjust pricing methodologies, test changes, and push updates to production with support from the Data Office. In under three months, over three-quarters of the team started contributing to full model development.
The team has also built user-facing trading applications using Databricks Apps and native dashboards. Core workflows including models, business logic, and dashboards run entirely within Databricks, with no separate BI tool required.
This unified approach reduces integration overhead and keeps business users close to the data that drives revenue decisions. Adoption is so successful that now most business users in the Trading department are actively building in Databricks.
For EnergyAustralia, the transformation is ultimately about speed and empowerment. The organisation can deliver trusted data to the right teams when it’s needed, enabling faster decisions in a highly volatile and dynamic market. As Dhivian Govender, Head of Data and AI, explains, “the goal is trusted information flowing naturally to where it's needed and when it's needed with business teams empowered to act on it.”
Federation as an operating model and the need for cost transparency
As EnergyAustralia moves towards a more federated data and AI operating model, cost visibility becomes essential. When domains own and access their own data, the organisation needs a clear line of sight into how the platform is being used and what it costs.
Dhivian Govender described the shift, "we're moving towards a more federated operating model at EnergyAustralia, and Databricks is a key enabler of that shift. Our stakeholders can now access refreshed data every five minutes on a self-service basis; no requests, no queues. The Data Office's role becomes less about delivering data and more about making sure the right foundations, governance and quality are in place so teams can move with confidence."
With more teams operating independently on the platform, cost discipline must keep pace. Instead of fixed on-premises infrastructure, EnergyAustralia now reviews platform spend monthly and monitors usage daily. Teams can track compute by lab, workload, and even individual queries.
Angela Merlino explained, "the granularity we have around costs is a game-changer. We can track spend right down to the user and the query. That visibility makes it easy to spot where we need to optimise, and when we do, we're not just cutting costs, we're building a more robust and efficient data foundation."
The team can flag inefficient queries, optimise models, and set alerts for long-running compute. Cost conversations are tied directly to performance and model efficiency as opposed to a monthly invoice. The result is greater transparency, tighter control, and a platform that scales with both the market and the business.