Women in Data: Doing Good with Data

Last Published: Nov 22, 2021 |
Susan Wilson
Susan Wilson

GVP North America Solutions Leader 

We recently held our third semi-annual Women in Data Leadership Forum, and I couldn't be more excited to share the results of our panel discussion. I'm passionate about helping women and minorities get into data and leadership positions. It's a rewarding experience for me, and I see incredible talent throughout the world that we can raise up in our organizations.

We were joined by four powerful women in the industry, who spoke about their data journeys and their personal commitment to helping women advance their careers in data-related fields: Maryann Byrdek, CIO of Feeding America; Terry Callaghan, AVP, IT and gift recording at Rutgers University Foundation; Rupal Patel, Director, Data Governance at the American Medical Association (AMA); and Anuleka Ellan Saroja, Senior Consultant at Deloitte.


Building Connections by Modernizing Data Analytics 

Panelist Maryann Byrdek kicked things off by sharing how Feeding America is working to centralize all their data from across the country to a national database and data warehouse. “We have a very complex environment with 200 food banks across the country, massive warehouses that help serve the 60,000 food pantries and meal programs. We do it all in service of the 38 million food-insecure in this country, including 12 million children,” Maryann said.


The organization strives to make the best connections between donors, food banks, and food pantries in the most optimized way, so that nothing is wasted and that food recoveries are as cost effective and efficient as possible. To accomplish those goals, Feeding America is building data repositories. “In terms of our data warehouse, we're building a lot of analytics dashboards, providing them all the way down to the food banks so that they can work better. Then they can service their end customers most effectively. We started this journey about two years ago, and we're about 20% done in the amount of data out there that we can collect,” Maryann explained. 


Terry Callaghan of Rutgers University Foundation shared how her organization’s efforts are similar, in terms of matching donors and their interests to initiatives at the university. “We have about a million constituents. About half of those are alumni. We bring in data from 30 to 40 different data sources, and so it's really important for us to make sure that our data is clean, and accurate, and consistent,” Terry said. 


“What we're trying to do is find out as much as possible about our donors, our alumni, our ambassadors—what are their real passions? Somebody might have graduated from the School of Arts and Sciences with an English major, but the data will tell us that they really have a real passion for working with autism, because that has touched someone in their family,” Terry added. Through mining the data, and through analytics work, and through the foundation’s data quality program, they can find that information, and then can match that person with that initiative so that they would be able to be an ambassador for it or make a gift to that program.


The AMA is also leveraging data to make better decisions. According to Rupal Patel, to get an understanding of what data exists within the organization, the AMA went through the exercise of identifying their enterprise data assets, and then investigated more deeply, asking questions like, “How are systems and applications using the data? How can we define it? How is it transformed?”


“We went through the exercise of inventorying our data assets, and centralizing the data, working with those partners to adopt the knowledge capital that we have centralized. We're hoping that as the tool becomes more enriched, we will be able to optimize certain data assets in a certain fashion based on business outcomes and use,” said Rupal.


Terry agreed that a data inventory is an important first step in a data program. Rupal explained how it helped increase data literacy in the organization as well, helping people understand how the AMA data is segmented, and how the different divisions intertwine. Connectedness to the data, to the subject matter experts, and to the business outcomes are all critically important.


Ensuring AI Is Ethical

With so much data at their fingertips, leaders like MaryAnn, Terry, and Rupal are leveraging artificial intelligence (AI) for scale and business innovation. But with AI comes the need to ensure that it is ethical. Anuleka Ellan Saroja shared how Deloitte operationalizes AI ethics across people, processes and controls, and technology. “Our point of view is that organizations need clear roles and responsibilities,” Anu stated. This is especially true for those stakeholders whose daily effort is to monitor and drive AI ethics. This may mean establishing a role of chief AI ethics officer, or an AI ethics advisory group, or distributing the responsibility across existing leadership.


You also need to train employees across the organization, especially those that are working with AI, she continued. She advocates creating a common narrative and lexicon for AI ethics, so that every employee that is working in the field of AI will start to think about AI ethics in the same way. Processes and controls are also needed, to establish a repeatable and sustainable approach to AI development and use. These processes contain guardrails that will map the technological solution to the AI ethics principles that the organization has outlined.


“There are several agile technologies that allow you to assess and truly validate whether AI tools are behaving in line with AI ethics principles,” Anu added. Deloitte advises and helps clients build technological solutions that can mine troves of data that are feeding into AI models and reveal insights and trends. These insights can answer questions around data bias, and data privacy, the explainability of AI models, and so on.


Building the Talent Pipeline

A number of new professions are arising from these newer data and AI ethics capabilities. The panelists shared their ideas for how to help encourage people to pursue these careers and how to nurture those who are just starting out.


Maryann advised that you start early on to encourage young people to consider a career in data. For example, it helps to take away the stigma of computer science and data analytics, particularly for girls. “A lot of young girls don't want to be in those geek propeller classes. You can be a marketing or business major and go into data analytics at the same time,” she noted. Start at the high school level and be a coach and mentor, she recommends, pointing out that a company like Google is a cool place to work, but you need to have data analytic skills to be effective in a work environment such as that or in any work environment.


It starts early, Maryann says, and you can also have midcareer pivots. At Feeding America, for example, many people move into the IT data analytics department that don't have a computer science background and receive training to support them. Terry concurred. The Rutgers University Foundation data scientist started out in the prospect development team, adding that mid-career pivots are really important.


Rupal agreed that it’s very important to draw others into the conversation using data. “I think women do a really great job of bringing others into the conversation and we should take advantage of that skillset that we inherently have,” Rupal declared. She also emphasized that it’s important for women to not worry too much about what they don’t know. “Women have a tendency to you feel that they have to know exactly what they need to do to move to the next level.” To counter that notion, Rupal likes to paraphrase a Mark Twain quote: “The secret of getting ahead is getting started.” 


Anu noted that there are also a lot of risks. “Gartner has predicted that in 2022, 85% of the AI models will have erroneous results. This is mainly because of bias in data, and teams responsible for managing them,” she explained. “One of the simplest ways to address bias in AI is by engaging diverse teams throughout the lifecycle of AI. By diversity, I mean, diversity of thought in the teams designing AI, developing AI, and deploying AI.”


On International Women's Day this year, Deloitte started a LinkedIn Live monthly series called “Leading Conversations in AI.” The series showcases women who are doing some of the most innovative work in AI. Similarly, Rutgers has a “Women in IT” initiative, Terry stated. The team found that the more diversity you have on your teams, in terms of ethnicity, race, and gender, the more productive and creative your team is.


Next Steps

For more insights from leading women in data, check out these video interviews with chief data officers Wendy Batchelder of WMware and Sonya Crosby of Westpac New Zealand.

First Published: Nov 23, 2021