Imagine searching through historical documents like never before! The PERO project aims to create tools to improve accessibility of digitized historic documents based on methods of machine learning (neural networks), computer vision and language modelling.