This project aims to build narratives through structured representations. This can be seen as a two-step process: first constructing the knowledge graph that acts as the backend common memory, and second generating narratives from that knowledge graph. The project has so far focused on three core use cases. The historical data use case focuses on event representations and enables question-answering over grand historical events. For the social media use case, we constructed a knowledge graph from twitter content using metadata and frame analysis, and we project to analyze current claims made by users on social media. For the social science use case, we built a hypothesis-centric knowledge graph and used it for automated hypothesis generation. The goal of the project is to build human-centric, interpretable solutions to assist humans in their daily tasks, and combines both symbolic and sub symbolic techniques.