Currently our service provides an API to serve the purchase records. The purchase records are stored in SQL database. To simplify, when users want to get their recent purchase records, they make an API call. The API call will run a SQL query on the DB and return recent records to callers.
Recently the purchase records data are migrated to Data Lake and can access through Azure Databricks. I go through the tutorial of Azure Databricks and can run some SQL queries on a cluster to retrieve those recent purchase records, but that seems more of a batch processing manner.
Now I am migrating our service to use data lake as a data storage. What is a good practice to support the scenario, where users trigger an API call to run a SQL query and get the recent records immediately? Would a Databricks SQL warehouse suitable for this?