You are constructing a machine learning model for real-time anomaly detection in sensor data. To manage incoming requests, Pub/Sub will be utilized. The goal is to store the results for subsequent analytics and visualization. How should you configure the pipeline?