The COVID-19 pandemic is a showcase for a data-driven society. Hence, the German government was aiming to provide free access to COVID-19 data to all citizens. However, making the corresponding data accessible by non-experts is not easy due to local characteristics and time-dependent metrics (e.g., the data is only collected on district level). We present the Coronabot facilitating the access to German COVID-19 data capable of answering German and English questions. The component-based system is capable of understanding questions relating time & (even small) places in Germany (s.t., a citizen might ask for the infection numbers in a self defined range of past time). A core demand was high and verifiable quality of the semantic search functionality as citizens need to trust in the provided data. As our system is driven by RDF (that means all internal components interact with each other using RDF and SPARQL), we are enabled to provide a controllable interface which is providing solid and traceable results. Hence, this significantly raises the level of quality assurance compared to traditional implementation approaches, while allowing microbenchmarking of each component using SPARQL on the collected trace information that is represented by RDF.