big database

Hi,

i have like 50 tables in a database, my main idea is extracting specific data to make calculations from it but the problem is the data is spread in like 6 tables totally apart from each other, and other problem is i cant just mix those tables bcs they dont have a direct correlation to match their info, for example: i have the table user and the table career, the career just have an identification and career name and user just have their personal info and their identification so i cant prove wich career belongs to who.

what im asking for is wich can be a good practice to manipulate the tables (one table can easily reach 2m rows), do i have to create a new table through dax code? or just mix most of the tables i need or get the data through measures? i dont really know how to filter a lot of data from differents tables to get just one result (for example the user and his career) to later make calculations.

thx