About me...

I am Junjun Lu (Lukas), a researcher in networking and sustainable machine learning at Universitat Politècnica de Catalunya (UPC). I received my M.Sc. in Advanced Telecommunication Technologies from UPC in 2025.
I bring a strong industry background spanning telecom backbone, data center networks, and cloud/security operations. Over the past years, I have worked in engineering roles across large-scale production environments, most recently as a Senior Analyst at Accenture.

My current research explores how to make large-scale learning systems efficient, reliable, and sustainable. Specifically, I focus on carbon-aware federated learning, multi-objective client selection, and budget-constrained scheduling that jointly accounts for communication, computation, and idle/waiting costs. I am particularly interested in bridging theory and practice through reproducible experimentation, combining trace-driven evaluation, network/system modeling, and automation workflows.