5 Q’s for Beena Ammanath, Executive Director for Data Science at General Electric
The Center for Data Innovation spoke with Beena Ammanath, executive director for data science at General Electric. Ammanath discussed how data science skills can apply across a wide array of different domains and how companies can help close the gender gap in data-focused roles.
Joshua New: Typically executive-level positions in the world of big data have titles like chief data officer or chief analytics officer. Could you describe the focus of your role as “executive director for data science,” and why this is a meaningful distinction from a more traditional title?
Beena Ammanath: Unlike the traditional titles like chief information officer or chief technology officer, data executive roles and titles are still evolving. The chief data officer role is typically about being able to look within one’s organization and manage the company’s data as a corporate asset. The role initially gained a lot of traction in 2007 after the financial recession and owned the data strategy, governance, and data risk management for a company. Nowadays this role has evolved into chief analytics officer titles, with expanded responsibilities to drive new business value leveraging data and analytics. We are also seeing the evolution of chief digital officer as a new title. So, there are multiple titles in the data space for senior leadership roles.
My current role is really about taking the best practices that we have applied within our own data and analytics teams internally as a large industrial company and making it available as products on our Industrial Internet platform. These products, ranging from industrial grade analytical models to industrial datasets, can be used internally across the company but can also be used by any industrial company in the world, so that we can all move faster together to drive positive outcomes for our end customers.
New: You have experience across many domains, including manufacturing, telecom, financial services, and retail banking. Is the role of data science similar enough across these sectors that it is easy for you to make these transitions?
Ammanath: I am a computer scientist by training. I have always been in technology-focused roles, and I have been lucky to work in an area, within computer science, that is my strength—data. I have worn multiple hats as I have progressed in my career. The domain, whether it is manufacturing, or finance, or marketing, has changed and I have seen how the whole data technology space has evolved. Thirty years ago, it was all about transactional databases, then came the era of data warehouses, and now it’s all about big data and data science. Ultimately, it’s about looking at data in different ways and being able to look at all kinds of data to drive a range of positive outcomes.
I have worked at large corporations, midsize companies, and startups. Data can be a key differentiator for all organizations. In the past, we used data to see “what happened,” and now with predictive analytics and AI we are able to predict, “what will happen.” It’s an exciting time to be in the data arena.
I believe the technologies that we use change, but as long as you focus on the outcomes or results that a technology can drive and not so much on the actual technique or technology, it’s very easy to stay relevant. I have always been interested in technology but it’s the use of technology that fascinates me more than anything else.
New: How would you define the Industrial Internet? And what are some of the most innovative uses of data that you have seen in the Industrial Internet space?
Ammanath: The Industrial Internet refers to a collective network of intelligent machines working smartly together. It is about optimizing machines, reducing downtime, and how it all translates into business value for our customers.
In the past few years, we started to notice that some of our aircraft engines were beginning to require more frequent unscheduled maintenance. By looking at just the engine’s operating parameters, we could tell that there was a problem. By pulling in the internal operational data and external weather data together and leveraging fleet analytics, we were able to cluster the engines by operating environments. We learned that the hot and harsh environments in places like the Middle East and China clogged engines, causing them to heat up and reduce efficiency, driving the need for more maintenance. We learned that if the engines were washed more frequently, they stayed much healthier and increased the lifetime and efficiency of the engine. All of this is possible because we could use data across every engine we manufactured, across the world and cluster fleet data.
New: Do you see much difference between the Industrial Internet and the consumer Internet from a data science perspective?
Ammanath: For both the consumer Internet and the Industrial Internet, it’s all about creating analytics that can help improve value for the business or consumer. It is the degree to which those analytics are used for mission critical applications that you would see most differences.
From a consumer Internet perspective, analytics are primarily used for improving the customer’s online experience by displaying more targeted ads and content. For an online retailer to know in advance when the online shopping cart may crash and avoiding such an incident can strengthen customer loyalty. This is an important business outcome, but is not as mission critical as an algorithm built for the Industrial Internet, predicting when a certain jet engine part might fail to prevent unplanned downtime, resulting in flight delays.
New: You’re also the board director for ChickTech, a nonprofit that encourages women to pursue STEM careers. What can public or private sector organizations do to reduce the gender gap in data-focused roles?
Ammanath: I think the gender gap needs to be fundamentally addressed at two levels. We hear a lot about the pipeline issue and there is a lot of awareness about getting more girls to join STEM. There are a number of organizations working on addressing the pipeline gap. However, we need more dialog on the women who are already in STEM and how to retain them.
According to a Harvard Business Review report, 41% of highly qualified scientists, engineers, and technologists on the lower rungs of corporate career ladders are female, yet more than 54% of them drop out between their mid-to-late-thirties. I believe that there are two primary reasons for this kind of exodus. One—subtle unconscious biases that exist at almost all tech organizations. And two—a lack of family initiatives at public and private organizations. I have personally experienced both during different stages of my career.
Companies need to research the biases that prevent women from getting ahead—study their data to find out if, and why, women are consistently promoted less often than men, create an inclusive collaborative culture where women don’t stand as out as the odd one out. Implement systemic changes within the organization that encourage women to rejoin the workforce after the birth of a child. Mentor and coach your leadership team to realize the value of diversity. Diversity is critical for any organization’s ability to innovate and adapt in today’s fast changing environment.