Yes, your point sounds much more feasible.

Sorry to ask for more, but can you link me to some document that describes the implementation in bit more details. I understood the overall concept, but just want to understand it better.

To be precise, Let's say we have 1000 wallets and every wallet has two wallet addresses. That brings that total to 2000 wallet addresses that may receive BTC.

If I am understanding this correctly, I need to store all these 2000 address in DB #1 and all txID in DB #2. But what I still don't understand is that I will still have to make call to bitcoin-cli for all the txID with gettransaction to find the wallet address and then compare it with the DB #1.

Correct me if I am wrong. 

Thanks.