OpenAI introduces GABRIEL, a tool designed so you, a researcher or a curious person, can transform texts and images into numbers that actually help analyze social phenomena. Can you imagine turning piles of interviews, course syllabi, or reviews into data ready for statistics without spending months labeling by hand? That's exactly what it aims to make easier.
What is GABRIEL
GABRIEL is an open source toolkit that uses GPT models to turn unstructured qualitative data into quantitative measurements. It's aimed at economists, social scientists, and data scientists, but it's built to require little technical expertise. Instead of writing complex rules, you describe what you want to measure in everyday language—for example, "how family-friendly is this job ad?"—and GABRIEL scores each document consistently.
The central idea is simple: the richness of qualitative sources (interviews, photos, syllabi, social media) often gets left out of large studies because converting them to data is so laborious. GABRIEL wants to open that path so you can study larger scales with less repetitive effort.
