By Mark Sellors, Mango Solutions
Last week saw the return of O’Reilly’s Strata and Hadoop World conference to London. Strata is focused firmly on the ‘big data’ end of the data science spectrum and boasts a phenomenal line up of speakers, 2000 attendees, and some of the industry’s biggest players as sponsors and exhibitors.
The Mango team staff the stand
This year, the conference was held in the Excel Centre in London’s docklands, which has a considerable size advantage over last years venue, both in terms of accommodating all the speakers as well as providing enough room for a large exhibition hall and dining area. In particular the keynotes benefitted greatly from the new venue: the room used for the keynotes last year wasn’t nearly big enough and many attendees had to watch the keynotes over the live stream in a room at the other end of the venue. No such trouble this year though and the keynotes were excellent, despite a power outage resulting in an unscheduled break.
Like many conferences, including our own EARL Conference, Strata starts off with workshops. Covering a variety of data science topics, in this case particularly those relating to Hadoop and big data, the workshops are a great source of information and insight. We were lucky enough to be presenting a workshop on using R with big data, which apart from a brief loss of conference wi-fi (which is always a danger at these events!) went extremely well. The workshop attendees were a really nice bunch and the workshop itself, which is a modified version of a workshop originally developed by our good friends at RStudio, was well received.
On the evening of the first day, local user group meetings, including our own LondonR, were given opportunity to hold their own events using the conference facilities.
LondonR was really well received, with talks from Jo-Fai Chow of H2O, Rich Pugh of Mango, Gabor Csardi talking about his R-Hub project and Tareef Kawaf of RStudio. The audience were great and the talks sparked some really good questions. It was also lovely to see so many regulars and new faces alike who’d make the trek over to docklands specifically for the event. Feel free to check the LondonR website for more info and past talks.
During the conference proper, there were so many great talks scheduled that it was often hard to choose between them. One that stood out for me in particular was from Brian Hills, of the Data Lab Scotland, talking about their work in improving awareness and skills. Some of the other talks I saw included Apache Beam, Apache Spark, Google’s TensorFlow ML Library, Apache Arrow, Apache Drill, and a really interesting one on lessons learned whilst A/B testing at Prezi. There are so many talks, attendees were spoilt for choice.
The exhibitors included, the big three names in Hadoop distributions, Cloudera, Hortonworks and MapR, as well as companies like IBM ,Dell and Intel as well as and a huge number of smaller companies from across the world.
Overall, it’s a great conference, and just the opportunity to network with so many like-minded people over the course of a few days in London is invaluable, and I’d encourage anyone to check it out. If I have one criticism it that it could do with more European speakers and business use cases as many of the talks are, somewhat inevitably, given by people related to the projects themselves, which are obviously heavily skewed towards the US. It would be great to see more people from this side of the atlantic speaking and sharing real world use cases with the audience. Keep an eye out and when the call for papers for next year’s event is open, take the plunge and submit your talk!
So, who knows, maybe we’ll see you there next year. We’ll be the people giving away all the cuddly cats.
