Cloud Hosting

Businesses today are collecting more data than at any other time in history but finding the time of day to keep data quality high remains an issue, says Kitman Cheung, chief technology officer at IBM Data, AI, and Automation in the Asia Pacific (APAC). Cheung is referring to some of the roadblocks faced by the banking industry in the APAC region, and how data and AI can play a crucial role to help banks overcome them and crest the innovation wave.

The speed and pace of data pipelines is growing more rapidly than ever, too, with a greater variety of data and disparate database platforms both structured and unstructured data, making their way into data repositories. This burgeoning volume of data must be managed, and doing it well is not a trivial challenge.

Data challenges

Onboarding new data sets and data sources into a traditional data warehouse continues to be a challenge. [The result is that] data sets either remain outside or data get onboarded at a much slower pace, making it difficult for stakeholders to look at the entire spectrum of data in a cohesive way, said Cheung.

In the meantime, the demand for new data continues to increase in both volume and importance. As banks start demanding greater access to data and at a faster pace, the capability gap between business needs and the ability of IT to deliver will only grow.

A dearth of skillsets is another challenge that banks in the region face, says Cheung. It is hard to hire data scientists who understand the banking sector on one hand, even as there are many business users in banks with the smarts and business knowhow but lacking in data science knowledge to collaborate meaningfully. The result is that innovation slows down at a time where banks need it most to stay ahead.

We see quite a few projects which were delayed due to the unavailability of data. To address that, a lot of clients are now looking at enabling self-service access to data to eliminate some of the bottlenecks in data delivery and IT transformation.

Agility without being bogged down

A fixation on new technologies or solutions for its own sake can often exacerbate the situation. Cheung pointed to how organizations around the world once gravitated towards Hadoop as the Holy Grail to address all their data-centric challenges. Often, they do not solve the [data] problems that we are facing. Instead, it can create a different set of problems, he said.

This can include popular platforms such as the public cloud: You outsource a bunch of capabilities to a third-party vendor. There are economies of scale for sure, but if you are using the cloud for consistent workloads and for a very long time, it starts to get more expensive. You start to think about everything in terms of OPEX (Operating Expenditure), which sounds good in the first place, but it translates into a different set of problems.

Instead of a complete rip and replace strategy with the favored technology of the day, Cheung recommends that banks adopt a more holistic approach and focus on fundamentals such as open-source software, open standards, and interoperability. Pick the right tools for the job. Continue with tools that are working and think deeply about data governance from the start not as an afterthought.

With a holistic IT strategy, banks can create an innovative platform that allows them to stay focused on the business objectives of every project. Leverage technologies such as containerization, microservices, and APIs to plug in new capabilities and evolve the platform, suggests Cheung. When it comes to actual investments in technology, organizations might want to think in small pieces, investing in a variety of interoperable technologies rather than an inflexible, monolithic system.

A fully integrated platform

A holistic approach makes it possible to build a fully integrated platform to support data science and AI. Cheung notes that such a data and AI platform can play a key part in supporting the culture, process, and people within the organization and a meaningful conversation around the same view on data. This is the reason why enterprises are moving to data virtualization, establishing a central place where all datasets can be accessed.

Crucially, such an approach allows a compliance policy to be seamlessly and cohesively applied across the organization instead of a piecemeal approach. Changes or updates to the data policy can hence be easily enforced without having to delve into individual databases or tables, ensuring personally identifiable information (PII) and data assets stay secure.

In closing, Cheung cautioned that data and AI tools are merely enablers to innovation, and not the solution: Successful transformation and becoming more agile and innovative is a cultural and process change. The company needs to say: I want to become more agile in business, and I willleverage data and AI technology to achieve that.

Paul Mah is the editor of DSAITrends. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose. You can reach him at paulmah@cdotrends.com.

Image credit: iStockphoto/Jeff_Dotson

Read the rest here:

Harnessing Data and AI to Crest the Innovation Wave - CDOTrends