Listening to Dr. Vishal Sikka’s keynote this year for some reason, I recalled his 2008-ish keynote about the notion of “timeless software”. That “vision” was fairly abstract, but I did get where he was going with it. However, sitting in this year’s keynote at TechEd in Last Vegas, I began to question…..”Are we really talking about timeless software anymore, or is it really the idea of ‘timeless’ data?”…..and I say “timeless” not with the idea of being “not affected by the passage of time” but more “not bound by the notion of time” all together.
I think we are at that point now (tipping point?) or very VERY close. With “traditional” thinking of data in relation to time…..past, present, future….and those limitations, I look at it like this:
- Historical data? Based on past limitations, companies often had to make decisions about “which” data to save and for how long. After so many years pass in the “life” of data, companies faced the decision of either (1) archiving data off onto some other storage in a not so easily retrieved manner or (2) “junking”/deleting/destroying “old” data all together. I always imagined “old” data to be like that warehouse at the end of the Indiana Jones “Raiders of the Lost Ark” movie…data goes there to never be seen again. (haha) Now, storage is “cheap”. The cost limitation is all but gone. Data of all sorts for all history can be stored. Large amounts of captured data can be gathered and kept for as long as company really wants it around. Historical data no longer has a cut off point.
- Present? We can now not only capture data in “real time”, but we can also consume that (and large amounts of data) almost instantly (and it is getting faster). This was never possible before some of the technology achievements in only the past few years! For example, we do not have to take sales orders, then move them to some other servers to then run reports which might come days later from to identify trends or such that we can only act on well after the fact. In the past, this was an accepted way of doing business. Now, it almost seems laughable….as in “did we really use to do it that way?!?!” Now, we can take those sales orders and have them instantly available for our reporting….actually showing trending in real time if we like! Captured and consumed at once!
- Future? We are now able to predict future data in almost real time as well. Further still, we can use the data coming in against those predictions to further refine future predictions and make them even better! Again, this was never possible before. Predictive analysis, data scientists, mathematicians coming up with better pattern recognition algorithms, etc. is a HOT area of activity these days.
I think Thorsten Franz nailed it when he predicted the convergence of OLTP (transactional data..ie, that data processed in real time) and OLAP (analytical data…ie. that large data churned and reported on at a later time via analytics) data. Those lines are blurred now as it is. Those lines will be gone VERY soon.
But there is even more possible now than just this. With a lot of movement to cloud solutions, there is now the possibility for companies to compare their data against all other companies’ data in the cloud for further amazing analytic capabilities. Company A can compare their data to Company B and Company C if they are all on the cloud. Sounds crazy? Well, SuccessFactors offers it NOW. Companies can decided to opt in or not as well as what data they share. Think about that past just HR related data like SuccessFactors into all other areas. Pretty amazing!
These new capabilities will bring about (I think) BIG changes in software (not so “timeless” eh? hah) as we no longer need all these MANY separate solutions for past data, capturing current data, and studying data. Software will be engineered and implemented with the thought going into the initial design that the data is unbounded in those “old” ways of thinking…developers will just assume that they have large amounts of data (old/new/predicted) available at any time to do with as they wish almost instantly. To make this possible however, the software has to “live” on something that easily supports all of this. Something like say….a “platform”? Like maybe the HANA platform? (see what I did there? haha) These are definitely exciting times!
But as the saying goes…”With great power comes great responsibility”. (haha)