Smart things can produce *lots* of data - in real-time. Cortana Analytics provides the tools you need to make sense of this data and react quickly to changing conditions. We will use as an example streams of real-time data from a car (ODB2 fuel levels and usage), fuel prices and position information (GPS) fed through Stream Analytics into SQL Azure. A trained Azure ML model will recommend the optimal time to refuel based on fuel price history, geographical location and predicted fuel usage. We will discuss how well known 'predictive maintenance' techniques can now be extended far from their engineering roots and be applied in many more situations to benefit individuals and organisations.*Note: Example will be 'canned' so it works, but I would like to try to show data coming from a real car if that is possible.Attendees will learn how Event Hub, Stream Analytics, Data Factory and Machine Learning can be used together to create valuable solutions and encouraged to start to think about the implications of IoT and how they might interact with smart devices.