Descripción
The role combines hands-on evidence synthesis with the development and evaluation of AI-driven approaches to make evidence synthesis processes more efficient. You will work on manual evidence-synthesis tasks, while also using Python and/or R to build and validate machine learning, automation and AI-assisted workflows for searching, eligibility screening, and (meta)data extraction. Your work will also contribute to new meta-research on how, and under what conditions, automated and AI-based tools can speed up and improve evidence synthesis in the environmental field without compromising reliability, robustness or scientific quality. Our team spans many disciplines and we welcome candidates from different backgrounds. If you have strong research and programming skills, want to help connect sustainability science with policymaking and are keen to learn evidence synthesis methods, we would like to hear from you. Key tasks and responsibilities Conduct and support manual literature searching, screening and (meta)data extraction for evidence synthesis projects, working with large volumes of scientific publications and grey literature (including bibliographic information and textual data). Develop and maintain reproducible workflows in Python and/or R for screening, extraction, classification, quality-checking, cleaning and harmonization of textual and bibliographic data. Apply and evaluate machine learning, natural language processing and LLM-based methods for text classification, information extraction and document analysis. Validate automated and AI-generated outputs against human decisions, quantifying errors, uncertainty and potential biases. Document methods, decisions and validation procedures to support transparent and reproducible research. Work collaboratively with researchers across environmental, sustainability, climate, energy and social science. Who you are You are an early-career researcher with strong analytical and programming skills who enjoys working where