SynthSprint: A synthetic data exploration

We are inviting expressions of interest to take part in SynthSprint, a four-week data sprint exploring how synthetic data performs across different levels of fidelity and real-world use cases.

The aim is to generate insights that will inform a developing synthetic data utility framework to support more routine provision of synthetic forms of secure datasets.

Much work has been done to understand what makes synthetic data useful. However, there is limited practical evidence on how different characteristics and fidelity levels perform when tested against real analytical tasks. SynthSprint brings people together to explore, compare and reflect on synthetic data in a hands-on, collaborative way. The aim is to generate insights that will inform a developing synthetic data utility framework.

Expressions of interest are open until 31 January 2026.

Participants will:

• Explore synthetic datasets with different characteristics and fidelities (low, medium, high).
• Test what is and is not possible at each fidelity level across real-world use cases.
• Experience collaborative, cross-disciplinary working (or individual exploration if preferred).
• Contribute to a synthetic data utility framework and a shared report/conference paper, with the opportunity to co-author. 

 

Structure of SynthSprint

SynthSprint runs over four weeks, combining two in-person workshops (pre and post sprint) with flexible, self-paced data exploration. The total time commitment is around nine days. Travel costs for the two workshops will be covered, and food will be provided.

Weekly one-hour webinars (Weeks 2–4) will introduce each dataset and task, after which participants can work at their own pace.

Who should apply?

We welcome applications from:

• Researchers and analysts in academia, national and local government, NHS/health sector, industry, and charities.
• People who routinely work with real-world data, especially secure forms of data (e.g. health, social science, economics, statistics).
• Participants at all career stages, including:
  • PhD students and early career researchers.
  • Mid-career researchers and data scientists.
  • Senior researchers and established PIs.

We will review applications and select individuals to ensure we have a good mix of skills, sectors and career stages.

Prerequisites

To participate effectively, you should have:

• Working knowledge of Python and/or R and able to read, modify and run Jupyter notebooks.
• Experience with data analysis and statistical analysis.
• Willingness to collaborate with others and work in controlled-access environments.
• Ability to commit time over the four-week programme and two in-person workshops.

Experience with synthetic data or trusted research environments is helpful but not essential. We are keen to include people who are curious and motivated to learn.

How to register

To read the full details and register your interest before 31 January 2026:

Visit the Datacise Open Learning platform