We are processing a stream of web logs. Basically activities that users perform on website. For each activity they perform, we a separate activity delta table.
We are exploring what is the best way to do streaming ingest. We have a kafka stream setup where all the activities are ingested in following format. But depending on the activity, we need to decide the different target table for the event to store.
{
activity_name: "Purchased"
data: {
product: "Soap",
amount: 1200
}
}
Can you help with what is the best way to handle this scenario?