A new law in Germany facilitates sharing solar panels in multi-party residential buildings. An energy management service provider (either a person or a company) can now maintain the appliances and distribute the produced energy within the building. For that, they need software to read metering data, perform automation, and handle invoicing. Additionally, future smart grids will allow extending Demand Side Management (DSM) to private households by connecting to this software, stabilizing the energy grid through dynamic demand adjustment in smart homes. One key challenge for such a system is the heterogeneous landscape of smart objects and network protocols-a problem that existing open source home automation software has already solved through ecosystems of plugins. This thesis investigates whether one of these existing products can serve as a basis for an industry-grade, multi-tenant Software as a Service (SaaS) product that energy management service providers can sell to many buildings. We contribute by combining existing research to derive requirements and a system design for the new scenario, and implement the most critical components in a prototype. We start by reviewing existing scenarios, standards, architectures and best practices for Internet of Things (IOT), smart home, smart grid, multi-tenant architectures, SaaS, and DevOps from literature. Then, we derive requirements and create a system design. After that, we implement a residential energy management platform based on existing open source technologies. The platform can collect sensor data at the edge using the home automation software openHAB and send it to the cloud. It can be deployed with almost no manual interaction through terraform, using configuration and infrastructure as code, and an edge optimized version of Kubernetes for orchestration of workloads. Additionally, we give detailed designs, access control patterns, and data flow sequences, and integrate the most relevant parts into the platform to show how the system can process collected data and achieve multi-tenancy. Eventually, we can show with our results that using existing open source home automation software in the given scenario is possible, but running it at scale for many tenants requires significant maintenance effort. We conclude by presenting a decision matrix with key decision criteria that influence the feasibility of the presented approach and some alternatives. This matrix helps decide under which circumstances home automation software should be reused and when another approach is probably the better choice.