I feel like if you're using those tools to process raw data (and not the data produced by another job), it's not really "big" data. I would tend to define big data as data where there's so much volume that processing it in this way is not possible.

Which isn't to say this article isn't full of hype—it clearly is.