This course introduces social science researchers to the practical use of Large Language Models (LLMs) in their work. Designed for non-programmers, the course covers the full research pipeline from finding and collecting data to analyzing it with AI tools and validating results for publication. Students will learn prompt engineering, work with real datasets from media archives and human rights organizations, and gain basic “vibe coding” skills to build simple tools without prior programming experience. The course addresses the challenges of working with hard-to-access sources, censorship, and sensitive materials in authoritarian and transitional contexts. By the end, students will have completed an independent research project using LLM-assisted methods.