The Belle II experiment at the SuperKEKB accelerator (KEK, Tsukuba, Japan) collected its first $e^+e^-$ collision data in the spring 2019. The aim of accumulating a 50 times larger data sample than Belle at KEKB, a first generation $B$-Factory, presents substantial challenges to both the collider and the detector, requiring not only state-of-the-art hardware, but also modern software algorithms for tracking and alignment.
The broad physics program requires excellent performance of the vertex detector, which is composed of two layers of DEPFET pixels and four layers of double sided-strip sensors. In this contribution, an overview of the vertex detector of Belle II and our methods to ensure its optimal performance, are described, and the first results and experiences from the first physics run are presented.