Skip to main content

Graduate Business Analytics Club Presents Analytics for Good Panel Discussion

Executives reveal ways business analytics is helping to create a better world.
Analytics executives Julie Grantier, Aashwin Jain, and Matt Cox spoke at the Analytics for Good panel discussion at Scheller College of Business.

BAC Corporate Engagement Manager Sherri von Behren; Moderator Abhinav Navuluri; Graduate Business Analytics Club Co-President Michael Veal; AtlytiCS CEO Julie Grantier, North Highland Director Aashwin Jain, and Greenlink Analytics Founder and CEO Matt Cox.

The Graduate Business Analytics Club hosted an informative and timely panel to discuss how business analytics is increasingly being harnessed to address pressing global challenges--and change the world for the better.

Led by moderators Meghan Flanigan, MBA, and Abhinav Navuluri, MBA, the panel consisted of three executives—Greenlink Analytics Founder and CEO Matt Cox, (PhD '14, Energy and Environmental Policy and MA '09, Public Policy); ATLytiCS CEO Julie Grantier; and North Highland Director Aashwin Jain—who detailed various ways their companies use business analytics to combat economic, social, and environmental challenges and make a positive impact.

"Analytics are playing a large role in driving major decision-making to help improve the human condition,” said Michael Veal, co-president of the graduate Business Analytics Club. “Our panelists covered issues like: What areas are most likely to experience fire rescue? What are steps we can take to help with the energy crisis? What methods can we employ to best meet the needs of those facing homelessness? Thanks to our amazing panel, we were able to learn how skilled organizations, such as ATLytiCS, Greenlink Analytics, and consulting firms like North Highland, are channeling the power of analytics to help answer these questions."

Aditya Hedge, a master's student in Computational Data Analytics and a vice president on the Business Analytics Club's leadership board, also found the panel informative. 

“The panel discussion offered a refreshingly realistic perspective of what it means to be an analytics professional in the non-profit sector," said Hedgde.  "It went beyond the ideal to show the problems that non-profits face and the skills necessary to overcome them."

Insights and stories from the panelists included:

Analytics for Good Examples

Matt Cox: Our flagship tool Athenia, is a recurrent neural network (RNN) that analyzes how the grid is going to operate on an hourly basis for the next 25 to 30 years. It measures 250,000 data points on every power plant in the country and builds a forecast, which we use like a digital twin. It evaluates data points like: what happens if we add more energy efficiency or rooftop solar? How will the system rebalance if we take some plants offline? We used that to create Atlanta’s 100% clean energy plan.

We also used a machine learning model that could predict displacement in cities across the country and deployed that in Chicago, Atlanta, and San Francisco. The model demonstrated that the energy burden was a statistically significant driver of displacement in all 3 cities. It's the only indicator that was significant everywhere.

Aashwin Jain: Many of our economic pro bono projects are geared toward fostering economic stability. In Atlanta, we used to work with Refuge Coffee when they were quite new to help them set up their operations. Recently we've been working with some other nonprofits like the Food Bank and the Boys and Girls Club.

A lot of what we do is actually for our clients in the corporate area, and then we see how those advancements can be translated into opportunities or projects for the pro bono space. Pro bono work is usually geared towards operational efficiency or helping showcase the value or ROI because money is always tight.

Julie Grantier: An awful lot of our work is about operational efficiency. All organizations, corporations too, but particularly nonprofits, have limited resources. They can only serve so many people. We try to determine the best way for those companies to utilize those resources and make the biggest difference.

For example, we worked with the DeKalb Fire Department on how to distribute education and intervention funds, and we also helped create a fire risk model. Most fire risk models are focused on commercial properties and financial risk, but if you're in a very low-cost neighborhood, you don't qualify as high fire risk. Our model focused instead on more tangible human aspects, including residential properties.

Currently, we're finishing up a project with a community farmers market. They put the markets in MARTA stations to try and mitigate the impact of food deserts by allowing commuters to grab something at the train station on the way home. This involved analyzing rider data, foot traffic, and a lot of different factors to come up with models to predict which stations would work best, how to reach the most people, and the optimum times of day to be at each station because volunteers are limited.

What are the challenges?

Matt Cox: Data literacy is almost always a problem. We spend a lot of time educating our clients about the importance of data and the difference it can make.

Funding and time are also usually barriers. Government staffers are usually insanely busy. That makes everything very stop and go, and the project can take a long time.

Aashwin Jain: When you're working for a for-profit client, you'll leave them with something, and then when you come back six months later, there will have been some movement. In the nonprofit space, that is not a guarantee. You might build or create something that everyone appreciates. But, then no one really knows how to take the project forward. Having to start again the next time you work with the same firm can be disheartening. But I think it's really important to continue providing them with support and/or some connections to keep them going.

Julie Grantier: We work directly with nonprofits, and sometimes they have six people working there, and none of them are analytics professionals. They have important jobs, working directly with the public and there’s no time for them to learn analytics so communication can be tough.

In addition, we're not working with big data and we're not usually providing the data. We're looking at their data. It's not clean, it's not organized. It is pretty much the digital equivalent of a bunch of Post-it notes. Getting the data to the point where you can begin to do analytics is a huge challenge.

What should current students upskill in?

Matt Cox: There are two things I want to highlight.

One of these that's been core to the success of our organization is what we call our listening tours: talking to as many people as we can to understand what their needs are. Do we understand the problem they're dealing with? Our clients may not understand what they need to solve the problem, but they always have a pretty good idea of what the problem itself is and what they wish they could do. I think learning how to do that will set you apart from other people. It sounds really basic, but very few people are actually prepared to do that well.

The other one would be data storytelling. The ability to contextualize and tell the story that the data is presenting, that's what moves the needle.

Aashwin Jain: Figuring out how to use AI models and LLM models for client applications is key. If you can introduce those models and enable your clients to become more efficient, that’s a strong value add. For example, if you can help them cut down the time spent crunching data, they’ll have more time to focus on value and strategy. I think that's a good area to educate yourself and spend your time.

Julie Grantier: Everything's going to be different in five years, 10 years, 30 years. The biggest thing is to learn to learn—and you better like learning because if you're in this analytics field, it's moving fast.

 Spend your efforts learning the business problem underneath the data strategy. There are so many data projects that are never used and don't show their value, because somebody didn't spend the time to figure out if that particular data project really addresses the underlying business problem.

This website uses cookies. For more information review our Cookie Policy