Database Architecture Design - AWS/RDS/DynamoDB/Etc -- 2

Help us decide on the best database to use for our project and then design the most cost-effective and appropriate database schema based on application needs. We use AWS to host all of our services and prefer to continue using them for this database but ultimately want to use the best database for the job. This database will be used to store various pieces of user information with the key identifier usually being a users email address. The data inserted into the database comes from many different sources and contains different information associated with a user based on where we received the data Similar services that store the same type of information include [login to view URL] and [login to view URL] The following link described how one service did it with Azure. [login to view URL] <b>Database Requirements:</b> - Ability to scale to billions of rows - Quickly return results for any given search type. - Search based on wildcards - Results returned in milliseconds. - Store all information related to the email address. - Store lots of unstructured data related to the email address. - Allow for detailed reporting of certain data points. (What is the most common password, what is the most common email domain, etc) - Ability to edit the data after it has been inserted - Store where the data came from. <b>Normal searches on the database are usually based on one or more inputs:</b> 1. Email Address 2. Email domain 3. IP address 4. First and Last name 5. Phone number 6. Username <b>Deliverables:</b> 1. Suggested Database 2. Suggested Database Schema 3. Guidance on how we deal with multiple records based on an individuals email address? Do we only allow one "database row" for all data associated with an email addess or individual rows for each new dataset? 20% - 30%+ of most emails are duplicates 4. Help setting up the initial database on AWS.