We present the BrainScaleS-2 mobile system as a compact analog inference engine based on the BrainScaleS-2 ASIC and demonstrate its capabilities at classifying a medical electrocardiogram dataset. The analog network core of the ASIC is utilized to perform the multiply-accumulate operations of a convolutional deep neural network. We measure a total energy consumption of 192uJ for the ASIC and achieve a classification time of 276us per electrocardiographic patient sample. Patients with atrial fibrillation are correctly identified with a detection rate of 93.7(7)% at 14.0(10)% false positives. The system is directly applicable to edge inference applications due to its small size, power envelope and flexible I/O capabilities. Possible future applications can furthermore combine conventional machine learning layers with online-learning in spiking neural networks on a single BrainScaleS-2 ASIC. The system has successfully participated and proven to operate reliably in the independently judged competition "Pilotinnovationswettbewerb 'Energieeffizientes KI-System'" of the German Federal Ministry of Education and Research (BMBF).