.jpg)
In an energy market dominated by traditional giants, the Danish startup Float has shown that it is possible to operate as an electricity retailer with a team of just three people, relying on artificial intelligence and an extremely efficient data architecture. Its value proposition focuses on offering households a real-time breakdown of consumption by appliance, something that until now seemed reserved for large-scale technological deployments. The key to its success lies not only in the machine learning model that classifies loads, but in an infrastructure decision that allows storing and processing millions of readings per second without costs skyrocketing: a 99.3% compression in its time-series database.
For any company looking to scale with limited resources, the choice of technological tools is critical. Float needed a database that could support massive data ingestion at 1 Hz per household, complex SQL queries, and continuous aggregations for real-time billing. After evaluating multiple options, they opted for TimescaleDB on Tiger Cloud, achieving compression that reduces storage volume to less than 1% of the original size. This allows them to retain all raw data to train their artificial intelligence models without worrying about storage costs. The lesson for other companies is clear: a well-designed architecture, combining AI for businesses with optimized storage, can make business models viable that would otherwise be unfeasible.
Behind this achievement lies a pragmatic approach to technological development. Float built its own hardware module to capture data from the smart meter, a signal processing pipeline, and a neural network for appliance classification. All of this is orchestrated from a cloud platform that combines Azure IoT Hub and Google Cloud. This type of hybrid cloud integration is increasingly common, and companies like Q2BSTUDIO offer custom applications that allow startups and corporations to implement similar solutions without having to manage the entire infrastructure from scratch. The ability to develop custom software for each need—from data ingestion to the user interface—is what sets apart projects that truly scale.
The Float case also illustrates how artificial intelligence can be integrated into the daily operations of a regulated company. Its system not only classifies consumption but uses AI agents to interact with customers, send proactive notifications, and automate billing and customer service processes. This allows a team of three people to serve hundreds of households, something unthinkable without an intelligent automation layer. In a context where cybersecurity and data privacy are critical, especially when handling sensitive energy information, having robust protection measures is essential. Q2BSTUDIO integrates cybersecurity into all its solutions, ensuring that data and systems are protected against threats.
The underlying cloud infrastructure is another fundamental pillar. Float uses AWS and Azure cloud services in combination, leveraging the best of each platform. Real-time data management, continuous aggregations, and threshold-based alerts run without manual intervention thanks to the database capabilities. For companies that need to transform large volumes of data into actionable information, business intelligence services like Power BI can visualize these consumption patterns and aid in decision-making. The combination of Power BI with real-time data allows energy managers and users themselves to understand their electrical footprint instantly.
Float's vision goes beyond household efficiency. Its goal is to turn the home into an active partner of the electrical grid, enabling smarter energy trading on the spot market. To do this, they need to continue scaling their platform while maintaining the same operational efficiency. This is where the concept of AI agents becomes even more relevant: autonomous systems capable of deciding when to charge an electric vehicle or when to feed energy back into the grid, based on prices and consumption predictions. These types of functionalities require continuous development and a flexible architecture, only possible with custom applications designed to evolve with the business.
In summary, the Float case demonstrates that innovation in the energy sector is not reserved for large corporations. With the right tools—cloud, artificial intelligence, specialized databases, and a focus on automation—a small team can compete and offer a superior service. At Q2BSTUDIO, we accompany companies on this path, providing custom software development, integration of AWS and Azure cloud services, and implementation of artificial intelligence solutions that make scaling with limited resources possible. Technology is the enabler; the right vision and architecture make the difference.
