In software development process, initial activities can involve requirements elicitation, modeling and analysis (verification and validation). The use of natural language in the register of the requirements facilitates the communication among stakeholders, besides offering possibilities to customers and users to validate requirements without extra knowledge. On the other hand, in the current global economy, software development for teams geographically distributed is becoming a rule. In this scenario, requirements verification and validation for medium or high complexity software can involve the treatment of hundreds or even thousand requirements. With this complexity order it is important to provide computational support for the software engineer execute quality activities. In this work we propose a strategy which combines natural language processing (NLP) techniques and software agents to support analysis activities. We have generated textual or graphical visions from groups of related requirements; visions help completeness analysis, identification of duplicities and dependences among requirements. We use content analysis techniques to support the identification of omissions in nonfunctional requirements. Also, we propose a strategy to construct the lexicon, using NLP techniques. We use software agents to implement web services that incorporate the related strategies, and also agents to act as personal assistants for stakeholders of the software project.