In large corporations, requirements are often stored in several locations and formats. Projects may use different tools, there may be several systems that are not directly integrated and it is likely that many people are working with requirements across different departments. Therefore, there is a need to collect these requirements to a single location and format. The main purpose of this thesis was to investigate how a general purpose tool using simple lunguistical rules would behave when used on texts in different formats, from different parts of the requirement process. The framework extracts requirements from texts, regardless of format, and formats them according to the users’ needs. The framework was developed using design research and was evaluated through a study at Volvo Car Corporation (VCC), where it analyzed three types of requirement documents: a legal text, a structured requirement document and an unstructured requirement document. The results of the study were then evaluated by the authors in cooperation with a representative from VCC. The results show that it is possible to use an automated, general purpose framework for this purpose with good accuracy. In comparison with current state-ofthe-art requirement extraction algorithms, the results of this thesis regarding the accuracy of finding requirements are equally accurate, with potential for improvement. The framework can be used by practitioners to collect requirements from different sources and migrate them to a given format.