Serious analytics takes a lot of time, cost and effort. One of the oft repeated complaints one hears is that by the time the reports are ready, the opportunity or the problem has ceased to exist. How can we make enterprise analytics faster, and cheaper?
In her HBR article titled “8 tech trends to watch in 2016” Amy Webb argues that Bots could be a big trend for 2016.
All of us have heard of software Bots- software that run over the internet, doing repetitive tasks. Over time they have come to acquire some negative meaning, mostly used in the context of social media, where they have stuffed everything from facebook likes to online polls. The slackbot is another famous example of a software bot.
By definition, Bots do largely repetitive tasks, and an enterprise has a ton of them. Therefore, the enterprise should be a natural playing ground for Bots. But that is yet to happen. What are the chances that Bots will become a major trend in enterprise IT in the coming year or the one ahead? Let us first examine where all can software bots benefit the enterprise.
One obvious area seems to be big data and analytics.
I am talking of something much like the SETI@Home program or the Zooniverse project, where large volumes of data are analyzed after breaking them into small lots. In the case of SETI@Home, software sitting on millions of PCs download data from a central pool, analyze it and post it back. In the case of Zooniverse, the data never leaves the central pool. Users log in and classify individual photographs.
What if bots were to hunt through large pools of enterprise data, looking for specific patterns? And raise alarms or inform the concerned employees when they find them? Would it be possible to bring down the compute cost of big data analytics, likeSETI@Home is doing?? Alternately would it be possible to speed up discovery, like Zooniverse is doing?
I do realize that all analytics problems cannot be solved this way. But there is a huge chunk that can be.
