Nuclear Energy Agency (NEA) – NEA coding competition gives students hands-on experience for practical AI-nuclear applications
Publish Date: 2026-06-26 07:47:00
Source Domain: www.oecd-nea.org
Winners of the NEA coding competition, Syrra Team 1
The OECD Nuclear Energy Agency (NEA) organised a coding competition that challenged participants to translate messy, unstructured construction data into machine readable data. In total, 40 teams participated with the winners flying to Jeju Island, Korea to join the NEA International Workshop on Artificial Intelligence for Nuclear Energy.
The rise of artificial intelligence (AI) comes as governments and industry pursue ambitious nuclear new build programmes to triple nuclear capacity by 2050. AI is being explored as a means to improve decision making across all aspects of nuclear power plants including in the management of construction risks, one of priorities for the sector. The NEA coding competition provided an opportunity to engage with and connect an exciting wave of young AI talent to the nuclear energy sector.
Competition design
The coding competition was designed to provide hands-on experience to students with a real-world problem that had practical applications in nuclear energy sector. The NEA chose risk registers, a widely used tool for managing risk in construction projects, as the focus of the competition. In practice, risk registers come in many different formats, often with unstructured information that is difficult for computers to read and understand. Participants were therefore tasked with automating the conversion of these registers into structured, machine-readable data. The winning approach included large language models (LLMs) and prompt engineering along with strong coding implementation to produce outputs that closely matched an expert-developed gold standard.
Competition outcomes
The NEA received high quality project submissions across the board, with participants developing a range of innovative approaches to the challenge. Among the 40 teams, Syrra Team 1, consisting of students from…
