AI Advantage Is Directly Proportional to Data Democratization — Mitsubishi Electric Trane US Director of Data Governance

(US & Canada) Sarang Bapat, Director of Data Governance, Mitsubishi Electric Trane US, speaks with Sue Pittacora, Chief Strategy Officer of Wavicle Data Solutions, in a video interview about the value of data governance through economies of scale, scope, and learning, and the future of data governance.

Mitsubishi Electric Trane HVAC US (METUS) is a leading provider of ductless and VRF systems in the U.S. and Latin America.

At the outset, Bapat demonstrates the value of data governance in work through a framework. He states that if a data governance program is carried out with a project mindset, then after a while, the excitement fizzes out.

Elaborating on the framework, first, Bapat ties data governance to economies of scale. For instance, if there is a production of 200 units, each is sold for $5, but if one produces 200,000 of those, then the cost per unit drops to 50 cents. Similarly, if data governance is done at scale, the time taken to search for the data needed to create a digital product goes down significantly.

With the data governance platform, when one types the requirement, it produces elaborate data sets with specifications and ratings. After choosing the required data set, it goes to the owner for approval.

Therefore, it reduces the search time of the data and, in turn, cuts down on the production cost of the digital product, says Bapat. He reflects that if the governance platform is at scale, one can find information at their fingertips.

In terms of metrics, Bapat stresses how many economies of scale systems are there on the platform. As a data governance practitioner, he notes that if the number of systems keeps increasing over the years, scaling will occur.

Next, Bapat mentions economies of scope and takes the instance of the re-usage of iPod technology in the iPhone, which shows how an organization increased price by adding an existing capability to a new product.

Translating that into data governance, Bapat states that building a data lineage is a regulatory requirement. However, if the data lineage is built accurately, it can be used numerous times to solve the data management problem.

Thus, by building lineage once, it can be used for both regulatory purposes and data management, says Bapat. His metric focuses on how many of these capabilities are built.

Then, Bapat discusses economies of learning and explains with an example of the Tide Detergent Company, which has research and development centers across continents. The reason lies in the difference in water quality in each geography, which requires them to create varied products. To do that, the research centers learn from each other how to better make the product.

Likewise, in data governance, says Bapat, as organizations scale and as data stewards get better, communities of practice are created, which are equivalent to economies of learning. Now, data stewards can discuss among themselves about enabling better outcomes with data. Summing up, Bapat affirms creating this concept for himself, using this framework, and sharing the same across other forums.

When asked about the future of data governance, he maintains that with the GenAI debate heating up, there has been an increased focus on having good-quality foundational data. This brings data governance to the forefront.

Further, unlike in the past, people are now asking more about how data governance can be tied to business outcomes, which is a massive change, says Bapat. With AI, he adds, analytics capabilities are in the hands of everyone, which, from a business perspective, hints at data democratization.

Furthermore, Bapat shares that CDOs have now understood the need to educate data stewards and the entire company about the data governance portal, the data assets, and what they mean in a business context.

With this knowledge of metadata, data ownership, data quality, and lineage, Bapat asserts that organizations can create better analytical products. He highlights that the ability to take advantage of AI is directly proportional to how good data democratization is.

According to Bapat, data democratization will be a big focus area for CDOs in 2024. He also discusses the changing role of data stewards, where it is not a casual role anymore. Bapat recalls meeting an executive whose title was Chief Data Steward.

In conclusion, Bapat affirms that CDOs would focus on tying data governance to business strategy, data democratization, data governance and GenAI, and data stewardship.

CDO Magazine appreciates Sarang Bapat for sharing his insights with our global community.