by Daniel Ives

AWS Certified Big Data Specialty

I'm in Las Vegas for AWS re:Invent 2016 and I've chosen to spend most of my brief time here taking three 170-minute exams for the new AWS beta-stage certifications. "Why???", I hear you ask. Well, it's like why people choose to climb Everest. Because it's there and they can. I will attempt to conquer these three exams and feed back my first impressions of each of them in turn. Remember, these exams are all beta, so the actual exams may differ significantly from what I discuss herein.

So here's my summary of AWS Certified Big Data Specialty (Beta).

What is it for?

According to the exam blueprint, this exam "validates technical skills and experience in designing and implementing AWS services to derive value from data". A major use case for the cloud is ease and relative inexpensiveness of extracting value from data, so this is another very useful exam.

What was it like?

Based on the previous two preview exams, I was expecting lots more questions. Only around a hundred on this one. I finished with about ten minutes to spare.

OK, but, you know, the structure?

It does exactly what it claims. There was some overlap with the security exam, to be expected as obviously security is an important part of your Big Data strategy.

Questions required a deep understanding of designing and using all of the data services available in AWS, from S3 through DynamoDB and on to EMR and RedShift. There were also questions on data visualisation, including but not limited to QuickSight (which I felt was a little harsh given that it's only been out of preview for a few weeks).

AWS Glue, announced at re:Invent, will also be a huge part of the Big Data story, so perhaps by the time the exam is launched, there will be some questions on that too.

As you'd expect from a Professional-level exam, you'd struggle to pass if you'd only done a bit of reading around the subject.

How hard was it?

The first couple of questions lulled me into a false sense of security. It seems to me that question order hasn't been randomised yet (for any of the betas), so those first few were bread-and-butter type questions. I was soon disabused of that foolish notion! As I've mentioned on the previous blogs in this series, you really need to know your apples on all the services which can be used to extract that elusive fourth "V" (Value) from your big datasets (the traditional three Vs being Volume, Variety and Velocity), but as this is a Professional level exam, you will also need to decide between several different candidate solutions to meet a specific business case. "Choose the right answer" becomes "choose the best answer" in many cases. Do I think I've passed? I'm really not sure; I took the betas in ascending order of the perceived level of difficulty for me, so I expected this exam to very tough. Having now taken all three, I'd say that Advanced Networking and Big Data are on a par, in other words I think I may have passed both (but equally may very well have failed both). But of course none of us will know until March when AWS Training and Certification have completed their statistical analysis

Final thoughts

As a major use case for many customers, borne out by the fact that there are already two "data" courses from AWS, achieving this certification will reassure your customers that you know how to design, build and use Big Data solutions in AWS.

Several courses are available which may help you to prepare for this exam - Big Data on AWS and Data Warehousing on AWS are obvious candidates - but there is no direct correlation, and those two courses alone wouldn't enable you to pass this exam.

Cheat sheet

Sorry if you came here for one of these, there isn't one. Work hard, play smart, and set up a Big Data pipeline!