Research Assistant

We are seeking to hire a Research Assistant (Pre-Doc) to join the research group of Dr Giuliano Casale in the Department of Computing and the f-inference project. f-inference aims to shift from a centralised, cloud-based deployment of Foundation Models (FMs) towards a Resource-driven Computing Continuum (RCC): a layered architecture that seamlessly spans Cloud, Edge and IoT tiers, deploying intelligence as close as possible to data sources and end users for a distributed, adaptive, sovereign, robust and energy-efficient paradigm of AI deployment.

Imperial College is consistently in the top 10 world university rankings with the Department of Computing ranked top of the 2021 UK .

This position will carry out research into the performance, cost and reliability modelling of distributed Foundation Model inference across the cloud continuum, spanning two main tasks. Firstly, the post-holder will develop simulation-based and analytical methods to evaluate adaptive caching policies for distributed FM architectures based on KV stores, quantifying the cost and latency of the caching infrastructure and its sensitivity to replacement, replication and time-to-live (TTL) configurations, and studying back pressure and other control mechanisms to mitigate hotspots.

Secondly, the post-holder will also contribute ICL’s cost and reliability modelling for distributed, layered systems, employing stochastic and multi-layer network models to characterise disruptions of expected latency and throughput, and recovery under failures, attacks or other reliability‑related events. The research will involve building and validating such models in simulation and experiments, but this is driven from the fundamental theoretical knowledge that underpins them, therefore contributions to knowledge come primarily from the modelling and AI‑systems elements of the project.

This position presents a novel opportunity to work alongside Dr Giuliano Casale and the f‑inference consortium in realising their full potential and planning for the future. f‑inference is a close collaboration of academics and researchers from across Europe as well as partners from industry.

• A first / Masters degree (or equivalent) in Computer Science or a closely related discipline.
• Proven research in performance, cost and reliability modelling of distributed and cloud/edge systems, or in AI / Foundation Model inference serving, with
  • Demonstrations of the ability to build proofs‑of‑concept of such systems
  • Practical Understanding of methods and techniques relevant to the role including:
    Performance & reliability modelling , LLM inference serving and cost optimisation, Resource scheduling, Libraries & Frameworks:
    Python, PyTorch, Num Py, Sci Py, Pandas, discrete‑event simulation, Anaconda, Jupyter, Systems evaluation: statistical analysis and performance metrics (e.g., latency, throughput, energy, accuracy).
• Strong software engineering skills to build simulation tools and prototypes will be an advantage.

You will have the opportunity to continue your career at a world‑leading institution. Imperial College is currently ranked #2 in the QS world university rankings with the Department of Computing ranked top of the 2021 UK .

You will receive a sector‑leading salary and remuneration package (including 39 days off a year) and a comprehensive early career development support package including 10 training and development days.

Please see job description for full list of requirements.

Full‑time, Fixed term to start October 2026 for 3 years with potential extension.

Please note that job descriptions are not exhaustive, and you may be asked to take on additional duties that align with the key responsibilities mentioned above.

We reserve the right to close the advert before the stated closing date, should we receive a high volume of applications. It is therefore advisable that you submit your application as early as possible to avoid disappointment.

Welcome to Imperial, a global top‑ten university where scientific imagination leads to world‑changing impact.

Join us and be part of…