Bhanu Suraj Malla is a Data Scientist at Amazon Inc. and a graduate researcher pursuing an MS in Computer Science at the Georgia Institute of Technology. His work bridges the gap between industry-scale operations—focusing on labor supply, demand, and compliance analytics—and academic research in time-series forecasting. His primary research interests include Bayesian hierarchical modeling, structural state-space formulations, and evaluating the empirical efficacy of statistical stationarity transformations. Bhanu Suraj is also passionate about open-source software development and building accessible analytics toolkits for the broader forecasting community.
How did you become a forecaster?
I stumbled into forecasting through the operational side of data science. Working on industry problems—like trying to predict labor supply, demand, and compliance at scale—made me realize just how much day-to-day efficiency relies on managing future uncertainty. Handling that kind of messy, real-world data sparked a deeper curiosity in me. It eventually drove me to pursue graduate research in Computer Science at Georgia Tech, where I could pivot from just using standard models to actually studying the underlying theory and structure of time-series analysis.
What areas of forecasting interest you?
My core interests lie in the structural and statistical foundations of time-series. Right now, I’m focusing a lot on Bayesian hierarchical modeling, state-space formulations, and probabilistic programming frameworks.
How has the International Journal of Forecasting influenced you?
Diving into IJF’s archives has been incredibly eye-opening as I transition deeper into time-series research. The journal has quickly become my benchmark for rigorous empirical standards. Seeing how contributors structure large-scale experiments and prioritize reproducible methodology has heavily influenced how I design my own controlled experiments in my own research.
What do you do in your free time?
Outside of my day job and graduate research, I give a lot of my free time back to the open-source community. I love building and maintaining targeted analytics tools for time-series data, such as https://make-it-certain.streamlit.app, a proof of concept in progress designed to help small businesses with demand planning. I’m also an active reviewer for the Journal of Open Source Software (JOSS), where I really enjoy vetting promising time-series applications that can bridge the gap between academic research and practical, real-world use.