Kaan Aykurt
PhD Candidate · AI-driven Networked Systems

About

I am Kaan, a fourth-year PhD student at the Chair of Communication Networks led by Prof. Wolfgang Kellerer. My research spans AI for Networking and Networking for AI. I work on reproducible evaluation frameworks for LLMs in networking and study communication as a bottleneck in large-scale AI/ML systems.

AI for Networking

  • Agentic LLM workflows for autonomous networks.
  • Benchmarking frameworks and reproducible evaluation.

Networking for AI

  • Communication bottlenecks in large-scale AI/ML training.
  • Evaluation of RDMA (RoCE, Infiniband) and standard Ethernet transports.

Selected Highlights

NetLLMBench
Benchmark framework for LLMs in network configuration tasks.
DML Network Traffic Analysis
Evaluation of distributed machine learning frameworks' network characteristics.