Over the past decades, numerous research and heritage institutions have taken the initiative to digitize their collections. This has led to a significant increase in the development and application of artificial intelligence (AI) techniques. This thesis focuses on the efficient use of AI to create and analyze various types of digital archives. Initially, the focus is on processing photo archives, where relevant people and objects are automatically recognized. The thesis then describes methods for geolocation and segmentation of topographic maps. After that, herbaria are processed, with a focus on automatic preprocessing and segmentation of plants. Finally, techniques for geolocating and analyzing social media are discussed. The results demonstrate how AI and data-driven methods are valuable for creating, analyzing, and managing digital archives. They save time and costs, and enrich the archives with additional metadata, significantly improving the accessibility of the collections.