Postdoctoral fellowship in line-rate machine learning at Networks Data Science Group

The Networks Data Science group at IMDEA Networks Institute has an opening for one Postdoctoral fellow in the area of applied machine learning (ML) to mobile network systems. The successful candidate will contribute to the ongoing activities of the NDS group on designing, implementing and demonstrating with real-world hardware efficient ML solutions for varied tasks (including telemetry, traffic classification, and anomaly/intrusion detection) in the programmable user plane of next-generation 6G networks [1-3].

In particular, the postdoctoral fellow will take a leading role in a large collaborative research project funded by the European Commission that involves major players in the telecommunication ecosystem, such as Telefonica, Telecom Italia, Nokia, or NEC. The fellow will oversee a team of two PhD students and visiting interns, and is expected to both to carry out independent, hands-on activities and to coordinate the work of the students towards experimental demonstrations of ML operation at line rate with production-grade equipment.

Publications

[1] A.T.-J. Akem, B. Bütün, M. Gucciardo, M. Fiore, Practical and General-Purpose Flow-Level Inference with Random Forests in Programmable Switches, IEEE Transactions on Networking, 33(5), Oct 2025

[2] B. Bütün, D. de Andrés Hernández, M. Gucciardo, M. Fiore, DUNE: Distributed Inference in the User Plane, IEEE INFOCOM 2025, London, UK, May 2025 - Best Paper Award

[3] A.T.-J. Akem, B. Bütün, M. Gucciardo, M. Fiore, Jewel: Resource-Efficient Joint Packet and Flow Level Inference in Programmable Switches, IEEE INFOCOM 2024, Vancouver, Canada, May 2024

[4] A.T.-J. Akem, M. Gucciardo, M. Fiore, Flowrest: Practical Flow-Level Inference in Programmable Switches with Random Forests, IEEE INFOCOM 2023, New York, USA, May 2023.

The position offers:

• Access to a cutting-edge 100-Gbps programmable network experimental platform featuring heterogeneous programmable hardware that includes Intel and NVIDIA ASICs, DPUs, SmartNICs
• The possibility of developing or honing PhD student supervising skills
• A chance to establish strong personal relationship with large industry partners
• An opportunity to develop soft skills in project management
• A vibrant, collaborative, multi-cultural and English-speaking environment
• The prospect of publishing at top-tier venues in networking
• The high quality of life of the region of Madrid, Spain, where we are based.

• A PhD in Computer Science or relevant field obtained by the joining date
• A solid research record with publications in top-tier venues
• Fluency in written and spoken English
• Enthusiasm for fundamental and systems research
• Proven competences in network programming (e.g., hands-on experience with p4, dpdk, pisa asics or smartnics) and, preferably, in applied machine learning (e.g., skills in designing and implementing ml models in keras, pytorch or tensorflow for practical use cases).