Academic conferences, vital for sharing research and fostering collaboration, have seen tremendous growth. But keeping track of all that scholarly data, from authors and papers to conference locations and topics, is a monumental task. Traditional methods struggle to keep up, leading to fragmented and unsustainable data silos. This is where Wikidata, the collaborative knowledge base, and Large Language Models (LLMs) step in. Researchers are now using LLMs to automatically extract crucial conference metadata from sources like websites and proceedings, enriching Wikidata with a wealth of structured information. Think acceptance rates, organizer roles, program committee members, best paper awards, keynotes, and even sponsors—all meticulously extracted and validated through a human-in-the-loop process. This project, focusing initially on Semantic Web conferences, demonstrates the power of LLMs to build a more comprehensive and sustainable scholarly resource within Wikidata. The results? Over 6,000 new entities added to Wikidata, and improved visualization tools like Scholia and Synia offering dynamic summaries of conference data through interactive charts, graphs, and timelines. Imagine exploring topic trends over time, visualizing co-author networks, or tracking conference participation demographics—it’s all possible thanks to this innovative approach. While challenges remain, such as ensuring data accuracy and preventing vandalism in the open-edit environment of Wikidata, this research paves the way for a more connected and accessible future for scholarly knowledge. The integration of LLMs offers a powerful solution to the challenges of metadata sustainability, promising a future where academic knowledge is more readily available and easily explored.