Perhaps this is not so surprising. Sheer volume is a solved problem, at least if you don't mind if operations take a while. Want to store more unstructured data? Fine, just add some nodes to your HDFS cluster. Want to do analytics on it? Go write some more MapReduce code, or if you're lucky, play with something a little smarter. Just don't expect results to come back in an instant.
Increasingly the people we talk to are looking to solve another problem: to turn the streams of data that their organisations generate (or that their customers produce) into meaningful information that can be used to support decision making. And since time machines are not widely available, the goal is increasingly to spot changes immediately and react before a problem escalates or before a competitor wins the business.
If you are not already a member of Big Data London you should seriously consider joining. The last meetup was held at IBM South Bank and we had some great speakers including:
- Chris Howard from IBM talking about Watson
- Chris Osborne from AltertMe with some great examples of visualising Big Data
- Mike Ferguson from Intelligent Business Strategies with some interesting views on the evolution of enterprise analytics
The next Meetup is co-located with O'Reilly's Strata conference (strataconf.com/strataeu) on October 1st at the Hilton Metropole on Edgware Road. Strata is well worth attending, if you can find the time. If you don't have the time for Strata, come and see us at the Meetup. You can sign up here: http://www.meetup.com/big-data-london/events/80788142/
And if you'd like to know more about what we're doing about the Fast Data problem and low latency real time analytics, please get in touch. We'd love to hear from you.