Data Skeptic

Student Spotlight: Aaron Payne, Data Analyst

26 min

Aaron Payne, an MBA student at Georgia Tech, discusses his transition into an analytics role at Chick-fil-A and his experiential learning project with Confama, a social services company in Colombia. He details how his team developed an interpretable ensemble forecasting model using SARIMAX and XGBoost to predict affiliated populations, emphasizing the importance of translating data into actionable insights for real-world impact. Payne also shares insights on navigating a career in analytics, the value of "decision science," and future trends like Generative AI.

Summarized by Podsumo

🎧 Listen 🎙️ Ask about this episode

✨ Key Takeaways

1

Aaron Payne's role at Chick-fil-A focuses on transforming raw data analytics into actionable insights and decision science, bridging the gap between data and strategic business outcomes.
2

He led a Georgia Tech project for Confama, a Colombian social services company, to forecast affiliated populations, directly impacting operational efficiency and the provision of essential benefits to people in need.
3

The team developed an interpretable ensemble model combining SARIMAX (with exogenous economic variables) and XGBoost to significantly improve forecast accuracy, while also addressing data cleaning challenges like language barriers and manual entry errors.
4

Instead of treating COVID-19 data as an outlier, the team used it as an indicator variable due to Colombia's history of similar economic disruptive events, making the model more robust for future volatility.
5

Aaron is excited about integrating Generative AI into workflows and aims to become a "Business Data Scientist" to effectively bridge advanced data science algorithms with critical business decisions.

💬 Notable Quotes

"Instead of a data science you have a decision science or instead of a data analyst our business analyst you have a insights analyst."
"Garbage in garbage out. So we want to make sure that the data that we're getting is not only robust enough to create good predictions, but also accurate and representative of reality."
"A smart man learns from his own mistakes and a wise man learns from others."