Help designing a simple but scalable DynamoDB

This is not a project per se, it's more an of questions for a project I am working on. I am designing a DynamoDB table that contains domain names (e.g: [login to view URL]) that were registered recently. This table must be updated once a day and usually grows at 100.000 to 200.000 records a day, grows at approximately 3 to 4 million records per month or up to 50 million records per year. I only have the following two attributes: * domain_name: This consists of the domain name, (e.g: [login to view URL], [login to view URL]). These values are MOSTLY unique however a domain name can expire and be registered again after one of a few years. * registration_date: This consists of the domain registration date (e.g: dd-mm-yyyy or 06-09-2020). These values are NOT unique. All domains inserted in the same day should have the same registration date. I need to conduct a lot of reading queries to this table and get results VERY FAST. I should be able to get all records between a date which can be millions. E.g: all records over the last 90 days. Questions to be answered. 1) What would be the best design strategy for this? OLAP vs OLTP? 2) Is the DynamoDB the right choice for a scalable and very fast database? 3) How this table should be designed? What must be the Primary/Sort key? 4) Since there a lot of users conducting the same results should DAX be advisable? 5) How can I build something can scales in performance but the AWS costs don't scale in the same proportion? Please feel free to share any other thoughts but please do not submit any answers before being accepted as the bidder, this will not make you the winning bidder.