EDITING BOARD
RO
EN
×
▼ BROWSE ISSUES ▼
Issue 34

The Big Data Revolution

Graham McEnroe
Content Manager at @ Ve Interactive
PROGRAMMING

Headquartered in London, Ve Interactive is a world-leading technology company whose platform has become an essential component for any online business to acquire, engage and convert more customers. After just 5 years, Ve has grown rapidly by successfully delivering products that give incremental value to our clients. Now with over 700+ staff, 26 offices worldwide and over 10k customers in 43 territories, Ve offers employees an entrepreneurial and global environment within which to grow and develop.

We plan to employ circa 1,000 people by the end of 2015, and to support this growth we are opening a new development center in Cluj. This being the latest addition to our portfolio of development centres, in London, Bilbao and Paris. This centre will assist us in not only developing our existing Apps, but innovating and developing new products to assist our customers in increasing their revenues.

  

Cyrille Vincey

Challenging the status quo in digital marketing, leading one of the top big data projects in the world, and the driving passion to become the next tech giant.

Ve Interactive Director of Data Cyrille Vincey gives his personal account of how Ve will take online marketing to the next level with big data and a unique vision.  

The turn of the year brought more than resolutions for Big Data and Data Visualisation company qunb, an acquisition from website conversion specialists, Ve Interactive. The goal is clear. To build a product which provides unrivalled business intelligence through real-time and retrospective exploration of online consumer trends, live analytics, and pinpoint predictive analytics.

In this interview, we speak with qunb's founder and now Ve's Director of Data Cyrille Vincey about just what he's determined to achieve with this fusion of talents, and what he needs to do it.

Can you please tell us a little about your role?

I run a team of data scientists, introducing storage data in a unified environment. When it comes to a mission statement, we have three primary objectives we would like to achieve.

To create new revenue streams based on data, augment Ve Interactive's existing suite of products using intelligence built on data, and develop new ways to acquire leads using data resources.

What excites you most about this project?

My passion comes from the people. Never have I met as much diversity and passion before joining Ve. There is also the tech challenge. We now have a vast playground of data in terms of size and resource. In regards to big data projects, this is a dream situation for any developer or engineer to be involved in.

What are the main challenges you face?

The pace with which the company has grown, even since my time, can make it difficult to keep track - in particular the rate of new product development and the unrelenting extension into new geographic markets.

While a challenge to keep up, this rapid expansion means the immense volume of potential data we can use grows also. The 10,000(approx.) Ve clients at present represents billions of pages and customer actions each day that can be analysed. Huge numbers. The sort that easily places this big data project in among the top 50 in the world. And with this broad scope of tech at our fingertips, it makes it impossible to stand still and the ideal place for constant learning. 

Big Data has become a buzzword - can add some clarity by giving me your definition?

I like to say that Big Data resurrects storage techniques which were in use during the 70s, but forgotten with the relational databases era of the 80s and 90s. The foundation of Big Data is simple: "If your data is too big to be stored on one single super expensive mega-server, just distribute the storage over a grid of small cheap servers."

Right after the definition, comes the part of Big Data that's exciting - well at least for data geeks. As your data sits over hundreds or thousands of servers, it's hardly ideal to move it somewhere else whenever you need to process it. Instead, you want to distribute your processing jobs over the storage nodes, and merge the distributed local results back into one place.

So in a nutshell, this is the cornerstone of any Big Data architecture - distributed storage plus distributed processing. Cherry on the cake, the open source world generates a wealth of complementary layers that you can plug onto this foundation, depending on the goal of your platform. Real-time processing, graph management, full text search, messaging, ad hoc analysis, to name a few. The open source community is contributing on amazingly powerful frameworks that now cover almost all the needs for data management over distributed architectures.

How do you see the use of Big Data develop over the next months and years?

We are still in the early education phase where we need to explain that Big Data is not the only answer to every data problem. Big Data is still a bit sacralized, but it becomes more and more mature, and more and more easy to implement, it shall be regarded as an alternative to relational databases in some specific cases. But not in all cases.

I'm probably more legitimately able to talk about the use of Big Data technologies at Ve. My apologies to the non-geeks, but let's do some name-dropping here. Our software stack is shaped after the lambda architecture. We are using Hadoop for distributed data storage, Kafka for message processing at scale, Spark for close-to-real-time data processing, MapReduce for batch data processing, Cassandra for the service layer, Titan and Giraph for a graph-based representation of our data, ElasticSearch to expose our knowledge graph to full-text search queries, plus a wealth of Machine Learning for classification and regression problems.

Long story short, we are using a 100% open-source software stack. And here comes the actual answer to your question: the next step for us at Ve is to open-source part of our proprietary techs, and as such contribute to the evolution of Big Data in general with our own R&D.

What effect has big data had on how brands and companies interact with their customers?

Correctly harnessing big data means companies have the technology to use any data for a multitude of different purposes. This technology means understanding patterns unidentifiable by humans. For example, detecting a correlation in the hours between when something occurs and levels of mobile activity.

It's this predictive capability that makes it so valuable, in as far as it's a way for brands to scale resources in response to anticipated trends. At the moment, most brands react to consumer behaviour online with a predetermined solution. With big data fuelled marketing, these interactions can be based on individual behavior - while using aggravated, anonymous information.

By these means, we want to improve the value proposition from Ve to its customers and also its customers' to their end user. 

Do you think Big Data is being used to its full potential at present?

We certainly have tools available, it's the data itself that's usually missing. That's what's so exciting about the proposition of joining Ve. We are now in the unique position of having access to immense volumes of data, which is complemented perfectly with the right talent and vision. 

How are Ve Interactive planning to revolutionise how big data is used?

We can contribute to the big data knowledge pool by participating a shedload of R&D into key areas. While it's important to continue nurturing advanced machine learning techniques, our primary focus is on real time responses.

We want to harness tech that works toward fulfilling this goal. To do so we need engineers experienced in open source development. Open source developers are the most passionate and dedicated in the world. Their work is bourn of irrepressible curiosity for their craft, which is something that can't be taught.

Why work for Ve?

We know for a fact that finding exciting big data development projects is difficult, except in the case of the long-standing tech giants. What we're doing will revolutionise web marketing with a vision unique to Ve on how to use data.

The most compelling reason to work here is that with Ve, you're challenging the present day status quo in digital. We don't want to be acquired by a giant, we want to become one. 

Conference

Sponsors

  • ntt data
  • 3PillarGlobal
  • Betfair
  • Telenav
  • Accenture
  • Siemens
  • Bosch
  • FlowTraders
  • MHP
  • Connatix
  • UIPatj
  • MetroSystems
  • MicroFocus
  • Colors in projects