NUFFIELD/Swiss National Research Foundation Research Engineer

Society; Oxford; Hybrid; GBP 50.000 - 68.000; Vollzeit

Benefits dieser Stelle

Research standing

Open-source visibility

41 days paid leave

USS pension

Free lunch and beverages

Career development support

Zusammenfassung

Nuffield College in Oxford is seeking a Research Engineer to lead the development of an open-source Python platform and to build scalable LLMS-enabled infrastructure for the Talking to Machines project. The role involves collaboration with the University of Lausanne and other partners, and requires strong Python and software engineering skills.

The successful candidate will work across data wrangling, machine learning, and optimisation tasks, with opportunities to present at workshops and

Qualifikationen

• Advanced degree in Computer Science, Engineering, Statistics or a related field.
• Excellent Python programming skills and strong software engineering practice.
• Experience building applications on top of LLMS and running large volumes of LLM inference at scale.
• Experience building and deploying cloud-based infrastructure.
• Ability to manage multiple projects and collaborate in cross-functional teams.
• Strong communication skills for technical and non-technical collaborators.

Aufgaben

• Lead development, release, and maintenance of the project’s open-source Python platform.
• Build and operate infrastructure to support large agent runs at reasonable cost.
• Develop pipelines turning public social media traces and survey data into persona representations for LLMs.
• Design and run evaluations to validate synthetic results; implement adaptive experimental designs.
• Work with data scientists, software engineers, and faculty to adapt code for needs of projects.
• Contribute to workshops and conference programmes to reach the research community.

Kenntnisse

Python programming

LLM applications

Cloud infrastructure

Data wrangling

Machine learning

Communication

Fine-tuning open-weight models

PyTorch

JAX/FLAX

TensorFlow 2.0

Hugging Face

Ausbildung

Advanced degree in Computer Science, Engineering, Statistics or related field

Tools

PyTorch

JAX/FLAX

TensorFlow 2.0

Hugging Face

Jobbeschreibung

Nuffield College and the Faculty of Business and Economics (HEC) at the University of Lausanne seek to recruit a Research Engineer to work with Professor Ray Duch (Nuffield College, Oxford) and Dr Sonja Vogt (HEC, University of Lausanne) on the “Talking to Machines” research project funded by the Swiss National Science Foundation.

The Project Engineer position is based at Nuffield College, working closely with the team of scholars in Oxford, the University of Lausanne, and collaborating institutions and partners worldwide. The core duties of the role include: leading the development, release, and maintenance of the project’s open-source Python platform; building and operating infrastructure that supports large agent runs at reasonable cost; building pipelines to turn public social media traces and survey, administrative, and census data into persona representations that condition LLM agents; designing and running evaluations to gauge the validity of synthetic results; developing LLM-guided adaptive experimental designs to enable the implementation of optimal treatments on-the-fly; implementing representation learning and text-as-data pipelines, employing learned representations of complex data for downstream (econometric) models; working closely with team members (data scientists, software engineers, and faculty) to adapt code used for data wrangling, machine learning, and optimisation to the specific needs of projects; and contributing to the workshops and conference programme through which the platform reaches the wider research community.

The successful candidate will hold an advanced degree in Computer Science, Engineering, Statistics or a related field. They should possess excellent Python programming skills and strong software engineering practice; demonstrated experience of building applications on top of LLMS, of running large volumes of LLM inference at scale, and of building and deploying cloud-based infrastructure; and proven capacity to manage work across several projects simultaneously, to learn new skills quickly, and to work well within collaborative and cross-functional teams. Strong communication skills for engaging with technical and non-technical collaborators are essential. Experience of fine-tuning open-weight models using parameter efficient methods, knowledge of a deep learning framework (PyTorch, JAX/FLAX or Tensorflow 2.0) and the Hugging Face ecosystem; experience of LLM observability and cost tracking; and familiarity with experimental and causal methods would be an advantage.

The post is full-time (35 hours per week) and fixed-term for two years, and available from 1 November 2026 or as soon as possible thereafter. The salary offered is £59,133 per annum. The Research Engineer will be employed by and primarily affiliated with Nuffield College where the successful candidate will receive supervision from and report to Professor Ray Duch.

What We Offer

• Research standing. Membership of the research community at Nuffield College — one of the world’s leading social science institutions — with the opportunity to co-author peer-reviewed publications and present at international workshops and conferences.
• Open-source visibility. The platform is released in the open; your work is public, citable, and portable.
• Generous leave. 41 days of paid leave per annum, inclusive of public holidays and College closure days.
• Pension. Membership of the Universities Superannuation Scheme (USS), with a substantial employer contribution.
• College life. Free lunch, coffee, and tea while on duty; membership of the College’s healthcare plan (single membership paid by the College); an Employee Assistance Programme; and a discounted travel-to-work scheme.
• Development. The College supports relevant training and career development, with financial support where appropriate.

The deadline for receipt of applications is Sunday 18th October 2026. Applications will be reviewed as they are received, and candidates may be invited for interview in advance of the closing date; early application is strongly encouraged.