Could you highlight the major findings of your technology trends report 2012?
There are five forces that are currently acting on businesses. These are analytics, social networks, mobility, the Cloud and cyber security. Each one of these is individually having an impact on businesses, although they have matured relatively recently and continue to evolve rapidly. Businesses have to leverage enterprise mobility, enterprise social networks and big data.
The success of any mobile platform depends upon how good the app store is. Mobile platforms and the usage of apps for business are especially widespread in SCM as traditionally the communication between dealers and suppliers was not structured in India.
What did you find with regard to big data analytics and mobility?
Analytics is a set of layers that a business has to grow from the bottom up. At the bottom is descriptive analytics, then there’s predictive and, at the top, prescriptive analytics. Descriptive would imply normal reporting as to how the business is going to do in the near future and how it’s done in the past. Not every business needs prescriptive analytics. However, this is all about structured data. Just like the above models are available for structured data, it is the same thing for unstructured data as well. In the case of unstructured data, it is mostly in the descriptive area. The industry has a lot of experience with structured data but it lacks the same with unstructured data.
If you look at how data is placed, there’s internal, external, structured and unstructured data. There may be valuable structured external data. E.g. for the insurance or pharmaceutical industry there will be regulations that are outside the organizations. One can apply this external structured data to internal unstructured data and gain value from it. There is a lot of big data in internal structured and external unstructured data. Organizations not only need to be aware of unstructured and structured, but also of external and internal dimensions that they need to deal with. Accordingly, tools and methods can be put into place. For big data these are still evolving. It is no longer a technology challenge. The challenge is going to be around getting the proper skill sets.
Presently, there is a lot of difficulty around decision-making when it comes to big data. One of the emerging trends is that of geospatial visualization. For an executive putting data on a map is an easy way to understand and navigate through it.
A few years back, we noted that CIOs had stopped investing in process automation and started investing in automating information. This was about automating ‘what do I need to do?’ as against ‘what do I need to know?’.
The relational structure is the most versatile when it comes to creating database drawings and it’s quite rapid with regard to transactions. The hyper cube is mathematically proven to be the most efficient shape for analytical transactions.
On the mobility front, it is all about building business apps for mobile platforms. The complexity is happening around managing these applications.
What is your take on columnar databases and in-memory computing and will it work in the Indian market?
The columnar database that SAP acquired through Sybase and in-memory computing are two separate items. But because they get bundled together they are confused. A columnar database is much faster than a row-wise database. One can compress the data. When you bring this into an in-memory scenario, the volume requirements are much lower. The cost of memory is going down. Databases are stored on a hard disk on a server and there is a lot of I/O taking place. Putting it in-memory makes it much faster to access. It will help bring in large volumes of data and much faster access to data which is what businesses need from a BI perspective.
What is your view on Cloud convergence?
The first generation of Cloud computing was essentially about capacity building and saving cost. It is a publicly valid business model for subscribing to Cloud services. The second generation is about vertical slices and it is what we call the capability Cloud. E.g. rather than Hadoop as a service it is the analytics Cloud; rather than SharePoint, it is collaboration as a Cloud; rather than storage as a service, it is backup and recovery as a Cloud. So there is lot of concentration now on building these slices of capability.
The challenge this year is the move from Cloud to Clouds. This is a hyper-hybrid Cloud wherein the integration of horizontal and vertical capabilities becomes important. Some businesses will use only the public Cloud and others only the private Cloud. However, the majority of businesses are going to go with the hybrid option within the enterprise with data driven integration. The PSU sector is a rapid adopter of the hyper-hybrid Cloud computing.