By Rob Starr, Big4.com Content Manager
Recent Capgemini research shows companies cutting across departments and adopting a centralized structure for Big Data and analytics have higher levels of success than their peers with decentralized models. In spite of this finding, 79% of organizations have not completely integrated their data sources. As a step toward a viable solution, Capgemini has announced a collaboration with SAP, Cloudera and Intel to develop a comprehensive framework to aggregate and synthesize scattered silos of structured and unstructured data. Anne-Laure Thieullent, Director of Big Data Europe/Global Big Data Solutions, Capgemini Group, filled us in on the bigger takeaways.
Why are these scattered silos of structured and unstructured data such an issue?
Silos of information exist in many organizations. It is a bellwether for how companies are functioning in terms of process, teams and units. It also offers a very good reflection of the history of this organization and how it was impacted by internal factors such as organic growth and external factors including new technologies. The other factor that influences the creation of data silos are how well – or not – an organization is inclined to work collaboratively cross BUs and LOBs, and this is heavily tied to the corporate culture. Finally, the power given to the CIO influences its ability to drive the pace of transformation and simplification across its technology infrastructure.
How is Capgemini addressing the need?
Traditional architectures have led us in the past 15 years to build enterprise data warehouses that would involve a complex process of designing an enterprise business into a data model, which would serve as the unique basis for all the reporting and analytics that a company would need to run its operations. Some of these projects have been extremely successful, but required sponsorship and direction from senior executives (e.g., a CIO enforcing that this new system will replace all departmental BI initiatives, and setting up the appropriate data governance to maintain that long term).
Businesses today are required to go faster. This need for more velocity requires more agile ways of working and collaborating around data. In that perspective, data warehouses are progressively becoming a bottle neck as, by design, they tend to enforce one and only one view of the data. As an example, the Revenue Assurance team at a Telco operator won’t have the same needs in terms of data quality than the marketing team, even though they would need to analyze the same elementary data – the Call Data Records. Trying to enforce one single view of the data for both teams won’t provide them the right answers, and ultimately they will choose to go their own way in terms of getting the answers they need.
To break away from those challenges, Capgemini invested very early in Big Data solutions, to shape, with its technology partners, the right architecture approach to put the technology at the
service of the business. With new approaches like Capgemini’s Business Data Lake that has now been adopted by EMC and the Insights Driven Operations solutions integrating SAP HANA and Cloudera, we aim to build for our customers new information landscapes that will truly deliver the agility they need in terms of getting to the right information, at the right time and at the right cost. These approaches use Hadoop-based business data lakes as a centralized platform to store all the data an organization needs to run their business in the most optimized way. On top of those data platforms, we work with our customers to determine the right analytical stacks to get the right tools in their hands and optimize their time to insights.
Who will benefit from the solution?
We designed the Insights Driven Operations solution integrating SAP HANA and Hadoop with
SAP, Cloudera and Intel, because Capgemini’s SAP customers needed a clear way to take advantage of the new Big Data solutions. We’re seeing the benefits of this integrated solution through different angles:
- Need to integrate insights generated on an existing Big Data platform, into the business processes, in real time when relevant
- Need to increase the scope of data that are available in the SAP applications
- For example, a retailer wants to integrate new and more complex data sets like web clickstream logs, to correlate customers’ online navigation patterns to their purchase behavior reflected in point-of-sales data.
- Another example would be manufacturers aiming to progressively move to connected devices with sensors, and need to monitor in real or near-real time and finely tune their production line or usage patterns.
What’s the next step?
We have built a comprehensive approach to guide our customers through a transformation journey via Innovation workshops where we guide them from an architecture perspective on how to transform the current state of their SAP application landscape to integrate new Big Data solutions. To start with, we have built demos that can bring the concept to life for our customers and serve as a basis to customize the solution and the approach to their most critical needs. To demonstrate tangible business value quickly, we set up a pilot project that will act as a foundation for the initiative moving forward, based on the first prioritized use cases.