Starting with mission

CORE-AIx starts from mission before methods.

The research path behind this lab has moved through several phases while keeping one constant theme: make advanced AI work under real constraints.

Before CORE-AIx was established, Prof. Shiqiang Wang’s early work focused on networking and edge computing, especially service placement, migration, and online decisions in dynamic systems. That then evolved into foundational contributions in federated learning and distributed optimization, where data, models, and compute are decentralized and efficiency is essential. Over time, the scope expanded to modern large-model workloads, including LLM training, routing, and inference design, with stronger emphasis on reliability and quality under practical limits.

This trajectory naturally leads to CORE-AIx: Collaborative, optimized, reliable, and efficient next generation AI at Exeter. The lab is designed as a focused home for integrating theory with systems practice, connecting collaboration, optimization, and reliability across the full AI lifecycle.

Some of the problems that we are currently interested in addressing include the following:

1. How to make agentic AI reliable and collaborative?
2. How does optimization theory interplay with large language models (LLMs) and generative AI?
3. How to make model training and inference efficient in distributed systems?