To truly enable grid flexibility, XENON must handle an immense amount of data, spanning from system-level programming on the edge to complex cloud applications. Wolfgang explains that the platform currently ingests roughly 100,000 requests per second from gridBoxes deployed globally. By focusing on shared-nothing architectures and high observability, the team catches sub-optimalities before they break, ensuring the scaling cloud infrastructure stays ahead of the growing load.
The data lake architecture at gridX is purpose-built to handle the massive volumes of time series data generated by assets like EV chargers and heat pumps. Chris details how his team has scaled these systems by a factor of 12 in just two years, providing the 15-minute historical data that forms the ‘backbone’ of our energy management solution. This high-quality data is what allows for advanced clean energy software features such as dynamic tariffs and load forecasting.
The team utilizes the ‘Inverse Conway Maneuver’ to structure teams around domain boundaries,like e-mobility or EMS,rather than technical silos. This approach reduces friction and allows everyone to own their features end-to-end, from the database to the API. The result is a massive boost in efficiency: we now deploy to production roughly 50 times per day, ensuring our smart charging and energy services stay agile.
When tackling legacy code refactoring, our engineers ditched rewriting in favor of incremental, parallel migrations. By running new code paths alongside old ones and comparing calculations in real-time, the team ensures that XENON remains stable. This ‘slow is fast’ philosophy, managed through a flexible Kanban workflow, allows the team to pivot instantly to production needs while maintaining a 99% self-healing rate.
“We’re having endpoints that are getting hit 100,000 times per second, 24/7, 365. In my world, that is a significant scale.”
“We've invested so much engineering into making it run robustly. The orchestration, the scheduling...it’s battle-tested in production, hardened. When an error occurs, it's self-healing. It can recover on its own.”
“This job is more about people than it is technical. Technical is only a small portion of what we do every day. I’m so fascinated by this topic: high-performing teams, creating high-performing teams.”
.svg){kind=icon}
.svg){kind=icon}