IBM Watson Data Platform: Make Data & Analytics Simple

Share this post:

Blogpost by Niko Tambuyser |

So you hire a bunch of young and ambitious data scientists to extract true value from your data. And then you find out that they really like open-source software, and that they all have their own tools they prefer using – making it really hard to work together and reach consistent analytic outcomes throughout your organization. Insights discovered in the data are not that easily transformed into value for your company. Does this sound familiar?

It’s what I hear more and more when I talk to companies which have taken on the big data challenge seriously to drive their business forward. I’ve got good news for them. Now there’s a data analytics platform that not only brings together data scientists, but also business users, data engineers and developers. All four user groups can collaborate smoothly on this platform, working together in projects on the same trusted data, with a user interface they feel confident in, and with all the tools and services they want – open source and all. I’m talking about the new IBM Watson Data Platform; you may already have heard about it.

One for all and all for one

This analytics platform was developed based entirely on design thinking and user experiences. Data engineers can connect to all kinds of data sources easily and safely – inside and outside the company, in the cloud or on the web, from the Internet of Things – and can shape and refine the data before making it available to their colleagues. At the same time, they can ensure that data governance rules are being respected. Business users’ lives have been simplified with guided and self-services analytics capabilities powered by cognitive computing. This includes asking questions in natural language, getting the answers in split seconds and having data visualized automatically in the right type of graph. Data scientists can choose notebooks and can code in the language they prefer, from Java and Python to Scala and R. Open-source integration is also part of the deal; consider, for example, visualizations, GitHub or Slack. As icing on the cake, developers can deploy the models built by the data scientists in batches and in real-time, or wrap them in APIs that can be built-in and used in new and existing applications.

Simple business example

Let me offer a simple business example, showing how the platform supports and shapes the entire analytics process in day-to-day operations. The account manager of a retail company sees the results of a product declining, and wants to do something about it. That’s where the data engineer comes in. He prepares the data and makes it available in the platform, so the account manager can easily ‘shop’ for data and apply the analytics services he wants using the self-service capabilities of his user experience. The account manager notices a correlation between the sales of the product in question and another article. Now it’s time for the data scientist to step in. He builds a model using any open-source tool and methods he wants, and finds that there is indeed a relation between the sales of the two products. He also discovers a correlation between age, gender and product groups. A campaign where a certain segment of the customers is offered a discount coupon on another product seems the solution for boosting sales again. The application developer applies this model to the retail firm’s web shop for cross-selling purposes.

More capabilities every day

The great thing about the IBM Watson Data Platform is not only that it serves the needs of different user groups in one integrated analytics environment. It can also be tailored to any company’s requirements, as it consists of various kinds of services and capabilities running on IBM’s Bluemix cloud development and services platform. The abilities of the analytics platform will be expanded continually: from predictive and prescriptive analytics to predict future outcomes and take the best possible next action, to machine learning helping to make the right choices. For example, data scientists won’t have to think much about the best data model to use – it will simply be suggested to them by applying machine learning techniques. Going forward IBM really sees the IBM Watson Data Platform as THE analytical platform.

Open source and cloud first

For open-source tools, the functionality of the data analytics platform will also continue to be expanded; IBM truly embraces open source and will follow developments in the market constantly. Currently already 27 data connectors are available on the platform and that number seems to grow rapidly. For the best possible performance, the IBM Watson Data Platform is powered by Apache Spark, the next-generation runtime environment for data. IBM considers Spark to be the ultimate operating system for data processing; in fact it will be built into all of its applications. As Spark is also available as a service on Bluemix, it offers a very scalable solution. You can get more processing power whenever you need it. IBM is taking a ‘cloud first’ approach with the analytics platform overall, and on-premises versions will be available in the future.

Need I say more? See for yourself why this analytics platform is the best for your data and your business, and try out our free
Data Science Experience Trial.

About the author

Niko Tambuyser is a Senior Data Architect at IBM and has a strong knowledge of Big Data & Analytics. Niko has more than 15 years of experience in the domain of data integration and data management. He has been advising many clients in Europe on data architectures and data warehousing, and he is a much appreciated speaker at seminars and events.