Skip to main content Skip to navigation

Joe Oakley

About Me

I'm a PhD researcher in Artificial Intelligence (AI) at the University of Warwick, originally from Liverpool. My research is generously supported by the Feuer International Scholarship in Artificial IntelligenceLink opens in a new window, and my supervisor is Prof. Hakan FerhatosmanogluLink opens in a new window. Before beginning my PhD in October 2020, I graduated with a First Class MEng in Computer Science, also at Warwick. I have been a member of the Computer Science PGR SSLCLink opens in a new window since the start of my PhD.

In my spare time, I enjoy playing the saxophone in various bands, both at Warwick and elsewhere.

Research

My research focuses on the design of scalable and cost-effective cloud-based AI systems. In particular, I am exploring the application of serverless computing to a variety of data analytics/ML workloads. These range from distributed deep neural network (DNN) inference to high-dimensional vector similarity search. I am also researching novel graph neural network-based approaches for spatio-temporal forecasting, particularly in the context of road networks (in collaboration with Transport for the West Midlands and other industrial partners).

Publications

Joe Oakley and Hakan Ferhatosmanoglu (2024) "FSD-Inference: Fully Serverless Distributed Inference with Scalable Cloud Communication". In: IEEE 40th International Conference on Data Engineering (ICDE) (to appear).

Joe Oakley, Chris Conlan, Gunduz Vehbi Demirci, Alexandros Sfyridis and Hakan Ferhatosmanoglu (2024) "Foresight Plus: Serverless Spatio-Temporal Traffic Forecasting". In: GeoInformatica Special Issue on Big Mobility Data Analytics (BMDA) 2023 (to appear).

Chris Conlan*, Joe Oakley*, Gunduz Vehbi Demirci, Alexandros Sfyridis and Hakan Ferhatosmanoglu (2023) "Real-Time Spatio-Temporal Forecasting with Dynamic Urban Event and Vehicle-Level Flow Information". In: 5th International Workshop on Big Mobility Data Analytics (BMDA) (* - equal contribution).

Aparajita Haldar, Shuang Wang, Gunduz Vehbi Demirci, Joe Oakley and Hakan Ferhatosmanoglu (2023) "Temporal Cascade Model for Analyzing Spread in Evolving Networks". In: ACM Transactions on Spatial Algorithms and Systems, Volume 9, Issue 2.

Awards

  • Best Presentation Award, WPCCS 2023 (Machine Learning track).
  • DCS Undergraduate SSLC Teaching Award 2020/21 (for CS346 Advanced Databases).
  • Finalist in Warwick Award for Teaching Excellence 2021.

Teaching

2020-21:

2021-22:

Contact

J dot Oakley at warwick dot ac dot uk

MB4.20

University of Warwick

Coventry, CV4 7AL