Data governance

Uber’s approach to data governance, CIO News, ET CIO

Data governance is an essential foundation for any organization that creates, manages or uses data. Data should be managed in a manner consistent with the organization’s data definition, data quality definition, and data governance framework. Inconsistent data can cause significant problems for the organization. These issues include increased operational costs, decreased customer satisfaction, and the risk of regulatory compliance issues.

Operating in 70 different countries and 10,000 different cities, Uber has implemented its data governance framework in an almost foolproof manner. Three things that have helped Uber do this are flexible engineering, keeping tabs on changing policies, and updating the framework as needed.

“When it comes to the global aspects, Uber operates in 70 different countries and they all have different governance frameworks depending on the laws they have. Data governance at Uber has always been given great importance. It is extremely important for It’s up to us to make sure that all of these laws and policies are followed, and we’ve tried to incorporate them into the software that we build,” said Manikandan Thangarathnam, senior director of engineering at Uber.

As he mentioned, Uber values ​​data governance as much as it does customer experience. Uber seems to have it all figured out as to what data they need, what needs to be stored and where.

While governance differs across cities and nations, Uber has built a common denominator and a strong pipeline for data governance. And on top of that, customizations and plugins are needed for each city and country.

The complexity though is building systems that can confirm this when a company operates in different cities with different policies. So, to solve this problem, Uber has built a platform where the essentials like privacy and security are taken care of. In addition, the company has configuration mechanisms in this platform through which it can customize each of these government laws according to the cities and countries in which it operates.

“For example, a city or country may have a rule that they don’t want the driver’s name exposed to the rider. So we have the possibility to turn on or off depending on the city, we do not need to create a different application for this. It’s the amount of engineering and flexibility we’ve built into our system,” Thangarathnam explained.

The first secret of data governance is to keep track of all the governance taking place in a particular city or country.

“We are making changes to systems in accordance with changing governance and policies so that our platform customers and partners are our policy complaint,” he added.

Uber’s four pillars of data governance are data collection, data storage, building intelligence and data science. Even though data is collected globally, the governance it must apply is different depending on the regions from which the data originates.

The data governance part continues to change frequently. Thus, the framework that one creates must be reworked again and again. But the data also continues to arrive every second. So the question really is how do you apply governance on top of the new data coming in. This is one of the interesting challenges that Uber is trying to solve today.

“Since day one at Uber, we have been working on our data governance framework. A lot of investment, people and skill went into creating what we have today. For companies now looking to expand into multiple geographies and across borders, they will need to invest heavily in governance,” he concluded.