I started writing stories on how companies and individuals deploy impactful AI applications and make money.
Here's the latest post - https://paradigmai.substack.com/p/how-a-bank-saved-500k-a-month-from
PS - I really hope this MLOps group had more activity on IH! Let come together!
Nice story. I'm wondering why use a relational database and not a specific for times series workloads?
In this case, the purpose of the DB was just to store the batch predictions and the actual data coming each day from the ATMs. So the relational DB was enough to handle that.