Curriculum Vitae

Kaan Aykurt
Download PDF
PhD Candidate · AI-driven Networked Systems
Technical University of Munich
Munich, Germany kaanaykurt@gmail.com linkedin.com/in/kaanaykurt kaanaykurt.me

Research Profile

Ph.D. candidate working at the intersection of AI for Networking and Networking for AI. I design agentic LLM workflows and benchmarking frameworks for autonomous networks, and study communication bottlenecks in large-scale AI/ML systems. My expertise includes reproducible benchmarking on hardware testbeds, transport layer analysis, and network modeling with Graph Neural Networks.

Experience

Chair of Communication Networks, Technical University of Munich
2022 – Now
Research and Teaching Assistant, Munich, Germany
  • Developing agentic workflows for network troubleshooting, implementing the Model Context Protocol (MCP) for tool calling to interact with network emulators.
  • Designed NetLLMBench, a reproducible benchmarking framework using Python and LangChain to evaluate LLM agents on network configuration tasks.
  • Built and operated a Kubernetes experimental cluster and developed a custom pod autoscaler (HyPA) to optimize resource provisioning.
  • Ranked 3rd in the GNNet2023 Challenge by designing Graph Neural Networks (GNNs) in PyTorch to predict latency on real-world datasets.
  • Conducted hardware testbed measurements of TCP/IP and QUIC to analyze congestion dynamics in high-throughput applications.
Chair of Communication Networks, Technical University of Munich
2020 – 2022
Working Student, Munich, Germany
  • Characterized traffic patterns of distributed training workloads (PyTorch DDP, MPI) to evaluate synchronization performance across RoCE and standard Ethernet.
  • Modeled TCP congestion behavior in reconfigurable data centers to predict flow completion times and quantify topology switching impact.
IBM
2018 – 2019
IT Specialist, Istanbul, Türkiye
  • Implemented automation workflows using Bash scripting and configured enterprise monitoring dashboards for operational insights.

Education

Technical University of Munich (TUM)
2022 – Now
Ph.D. Computer Engineering (Advisor: Prof. Dr.-Ing. Wolfgang Kellerer)
Technical University of Munich (TUM)
2019 – 2022
M.Sc. Communications Engineering
Koc University
2014 – 2019
B.Sc. Electrical & Electronics Engineering | B.A. Economics

Technical Skills

AI & ML Python, PyTorch, Distributed Fine-Tuning, Hugging Face (Accelerate, PEFT/LoRA), vLLM
Networking C++, RDMA (InfiniBand/RoCE), P4, eBPF, ns-3, Mininet, TCP/IP
Infrastructure Docker, Kubernetes, Helm, Linux, Bash, Git, GitLab CI

Selected Publications

Agent-based Enhancement of Small Language Models: A Lightweight Solution for Network Management
K. Aykurt, C. Dang, A. Blenk, W. Kellerer. Manuscript under review.
NetLLMBench: A Benchmark Framework for Large Language Models in Network Configuration Tasks
K. Aykurt, A. Blenk, W. Kellerer. NFV-SDN 2024.
When TCP Meets Reconfigurations: A Comprehensive Measurement Study
K. Aykurt, J. Zerwas, A. Blenk, W. Kellerer. IEEE TNSM 2023.
Network Traffic Characteristics of Machine Learning Frameworks Under the Microscope
J. Zerwas, K. Aykurt, S. Schmid, A. Blenk. CNSM 2021.