Several studies indicate that poor requirements practices, that result in incomplete or inaccurate requirements, poorly managed requirement changes, and missed requirements, are the most common factor in project failure. Possible solutions for better requirements definition include better requirements documentation, and requirements reuse. Natural Language Processing can help extracting and formatting requirements so that machine learning (ML) algorithms can then be used to recognize different types of requirements and their relationships. Thus, by classifying requirements, we increase the documentation quality, and potentially allow requirements reuse in future applications. Also, this approach would be beneficial to requirements documentation quality in startups’ contexts, where requirements are usually not specified systematically. Active Learning, a sub-field of Artificial Intelligence and ML, is an algorithm for scenarios with abundant unlabelled data but with high cost to manually label such data. This is invaluable, considering that requirements datasets can be huge and time-consuming to label by hand. In this paper, we use ML and Active Learning to classify the requirements of a given dataset. This approach can accelerate project development. By organizing the requirements into categories, developers can easily see what requirements were already implemented, and where they need to focus on the next step of development.