The question we get most often is "what should we use?". The honest answer is that the technology matters less than people expect, and three other factors matter more.
First: who maintains this a year from now? If it is your in-house team, pick what they already know, even if it is not the newest thing.
Second: how much data do you actually expect? A straightforward relational database carries most companies for years, and there is no reason to build complex infrastructure before you need it.
Third: how fast will requirements change? If the product is still finding its shape, choose tools that make change cheap rather than ones that lock a rigid structure in on day one.
