Creating real-time predictions through an inference pipeline