My previous blog touched upon the paradigm shift taking place in today’s increasingly distributed data landscape. As a natural extension of this transformation, I would now like to share some thoughts on how cognitive technologies are poised to power data marketplaces which have arisen in recent years. More so due to the integral role knowledge management now plays in the cognitive world with respect to training machines for business and data processes.
I believe that most 21st century enterprises have yet to fully tap the potential of cognitive technologies and have much more left to accomplish in the field. These technologies are far more powerful than the day-to-day data processes automation tools most organizations are aware of, and involve numerous innovations in machine-to-human or machine-to-machine interactions that can transform business outcomes.
Rise of Data Marketplaces
Just as oil was the most valuable fuel of the 20th century, many industry leaders believe that Data is the most valuable fuel for 21st century enterprises – but while oil may have its limits, data does not. Many organizations are already exploring ways to unify and speed up the way they use data in various business scenarios. As a result, we’ve seen the rapid emergence of Data Marketplace tools and platforms that seek to address this business requirement.
Essentially, data marketplaces are a “one stop shop” that businesses use for business requests, reports, and insights in an easy and direct manner. Within an enterprise, data marketplaces offer a simplified architecture that can easily collect, collate, organize and integrate, data from various sources to become a Unified Data platform. An effective Data Marketplace is able to structure the unstructured data across the organization, by creating a compatible data model across silos that standardizes data formats. This data is further emboldened through various enhancements such as data search interface and data visualization which allow for up-to-date, fast and easy projections and estimations of business scenario simulations, making it a valuable business tool.
This approach to accessing data allows businesses to utilize their company-wide data from anywhere through a single point of interaction. This saves time and effort, as data no longer needs to be sourced, selected, and interpreted for each business case. As a result, businesses are able to achieve cost savings and quick turnover timelines as issues of data replication and data movement are solved. However, as we continue to look at the changing digital data ecosystem, we can see that the potential of data marketplaces can be significantly enhanced when we leverage cognitive solutions to further extend their functionality.
For example, in one of my recent projects, I had the opportunity to work with one of the world’s leading market research companies to help realize their need for automated data delivery and processing. Together, we were able to truly unleash the business value of data by consolidating all their requirements into a graph database with a machine learning powered cognitive interface that could enhance search prioritization and quality matching. In this manner, cognitive solutions were leveraged to enhance quality requirements with the critical goal of ensuring that there was no mismatch in customer expectations vis-à-vis delivery.
Traditionally, most data operations have been done through traditional manual interventions that requires enterprises to spend an immense amount of workforce resources. However, cognitive solutions changes this as it can speed up the outcomes by simplifying tasks such as driving data governance, continuously updating metadata, handling knowledge management and most importantly, monitoring data operations.
Data Marketplaces in the Enterprise
Let’s consider for a moment the needs of the next-generation enterprise – speed, accuracy, and excellent customer experience. All these goals require companies to be quick to respond to changes, and develop well thought out business scenarios to leverage emerging opportunities.
Take for instance a major global retailer we work with who operates over 410 stores in 49 countries. The company has over 800 million annual customers, over 2 billion annual visitors to their stores, nearly 150 million app users, nearly 150,000 employees, and over USD 40 billion in sales. For such a company, the ability to integrate data from across diverse nations, customers, and people, doesn’t just end in the front-end but also extends across their value chain.
For them, ensuring a harmonious data ecosystem across geographies and sources is critical to being responsive, fast, and efficient in serving their customers across geographies with a high level of quality. Moreover, given their leading position in the industry, they also have access to data from numerous vendors, suppliers, and manufacturers that are part of the production value chain. Therefore, having active access to these disparate data sources, such as inventory data, finance data, supplier data, and customer data creates the ideal data pool needed to track lead times and efficiently manage their business plans over seasons, cycles, and different business scenarios.
In such a case, having access to a Data Marketplace is not only beneficial but also critical for sustained business growth and success. However, the challenge to being responsive at such a scale can prove daunting. While Data Marketplaces are useful in bringing data together, the final bottleneck remains invariably human. This is where AI and other cognitive technologies have played a big role and helped ensure rapid access to insights and intelligence.
Potential of AI in Data Marketplaces
Data Marketplaces in a cognitive world don’t have to simply be “one stop shops” but instead can act as deployed agents. These cognitive agents function like an engine that continues to handle data operations and governance including all administrative tasks to support daily needs such as information regarding consumers, business analysts, power users, executive leadership, casual users and most importantly, business process owners whose endeavor is to simplify processes to enable quick outcomes in a business agile world that is freed from a human worker’s knowledge and ability to make decisions.
Over the last few years, I’ve worked with my team to develop this form of cognitive tool that takes a modern approach to data exploration. We simply call it the HCL Data Bot that is integrated to our data fabric ecosystem in our solution to help enterprises as they scale digitally. This cognitive tool was designed to help organizations manage data operations, processes, and system performance in order to help contextualize business needs and enable monitoring of business KPIs. Consequently, the integration of this bot within data marketplaces has been able to help enterprises easily discover actionable business insights and help users sift through the hundreds of data variables and fields, while also being able to locate data and link it to the appropriate metadata for a more comprehensive dataset.
This type of holistic integration can help users acquire the most optimum data selection that is suited to their business scenario and leaves the need for manual searching in the past. In fact, with advanced AI and NLP based tools, data bots can assess a user’s data requirement history and proactively offer a more specific dimension of selections, saving time and effort. Users can also easily discover the data’s availability, and the bot can share various useful details such as the previous use-cases of a particular dataset, as well as the feedback surrounding its previous uses.
AI Behind the Platform
Throughout my career, I’ve always placed an immense importance on design thinking and knowledge management, both of which have been core pillars of digital transformation. This approach is essential as it is more capable of understanding user personas and their value journey, thereby enriching the value of the end solution. HCL’s Data bot can provide prescriptive suggestions for orchestrating new data sets as they become available on the platform, thereby making the process more proactive while also being reactive.
Cognitive intelligence tools can also be used to assess the core effectiveness of the data provided. By evaluating the degrees of success of past deployments, cognitive intelligence tools can help businesses quantify the data KPIs and assess its value in achieving business goals. This process enables businesses to effectively plan the correct data for the correct business problems, by knowing which scenarios offer the greatest ROI, and whether it leads to better business outcomes.
Moreover, cognitive intelligence can play a valuable role in assisting in the management of data processes at their very fundamental level. Similar to virtual personal assistants, advanced cognitive intelligence tools can help in the processes that surround data administration by monitoring data flows on the platform. These solutions can also ensure an automated performance and quality check of the data marketplace platform.
As I’ve said before – digital data ecosystems are undergoing a paradigm shift where we need to adopt a Data Fabric approach to make data more accessible and actionable across the enterprise. Moreover, as more and more organizations begin to adopt a fully digital approach, the challenges of digital at scale can only be tackled with the benefits of cognitive technologies. I’ve already had the opportunity to work on numerous implementations of this kind and am convinced that cognitive solutions offer all enterprises the correct combination of tools with which to unleash their business intelligence potential and blaze into the 21st century.