Some weeks ago, I posted the first part of the mini project that I was creating with a teammate about a initiative of my manager who encouraged us to learn new things in a different way.
This first article talks about how to ingest data in real time in Synapse, and it explains one of the multiple ways that we can achieve it.
Real-time ingestion with Event Hub and Synapse
Also, you can visit this other article that explains the other part of the architecture about batching ingestion.
End to end analytical pipeline demo with Synapse
Now the second part of the process is how can I transform my data and how can I work with to produce Machine Learning models so I can analyst cases.
This part is already completed in the repo: vasegovi/synapse-demo (github.com) first you have to complete Hand-On challenge to have the information for creating ML model and train data, you will get fun, I’m sure on that, so please go ahead and follow the exercise.
Some take away from that activity, there are different ways to integrate ML models in your azure synapse, you can create them directly from Azure Machine Learning workspace and then using it in synapse or the other side also. There are no right or wrong decisions about in which way you will work with that, a lot of things depend on the company policies, the architect decisions, but here you can find a complete example of how to use this Synapse, Machine Learning and Power BI artifacts where you can start.
And please share with me any feedback, questions, or comments that you might have.