Welcome!

I am a Visiting Assistant Professor of Econometrics and Statistics at the University of Chicago Booth School of Business and the founder and elected chair of the SPEC RG Predictive Data Analytics Working Group. Before joining Booth, I was an Assistant Professor of Computer Science at the Illinois Institute of Technology.

My research is in performance engineering for AI and cloud systems. My goal is to make the infrastructure behind modern AI efficient, predictable, and trustworthy, so that systems not only run fast but also behave reliably in terms of cost, latency, and correctness. To this end, I combine systems measurement and benchmarking with forecasting and machine learning.

My current work focuses on three directions:

• Efficient AI and Cloud Infrastructure: Understanding and optimizing how AI workloads run on container platforms, from GenAI inference on Kubernetes to container start-up and proactive auto-scaling based on workload forecasts.
• Reliable AI with Guarantees: Making LLM-based systems dependable, for example through LLM routing with distribution-free safety guarantees, fault-tolerant transformer inference, and structured LLM inference for high-stakes domains such as mental health.
• Performance Prediction and Observability: Predicting and explaining the performance of microservice applications, including their transient behavior under auto-scaling, composable observability queries, and LLM-based workflows for root cause analysis.

Beyond computing systems, I apply my expertise in time series analysis and machine learning in interdisciplinary collaborations, particularly in healthcare.