Natural language processing as work support in project tendering

When producing a tender, contractors manually analyze client requirements contained within many different text documents. The combination of requirements lead to crucial design decisions and every decision is related to costs and risks. This study explores the possibility of making the client requirement analysis in design-bid contracts automated to reduce the risk of conceptual design mistakes. The research approach chosen includes developing a work support tool based on natural language processing and analyzing its usefulness through a combination of surveys and a workshop for tendering specialist. The results show that applying digitalized working methods and using artificial intelligence in the tender phase can enable data-informed decision making and generate benchmarking and risk management opportunities. The study contributes to insights regarding automation and digitization possibilities in tender projects and how artificial intelligence tools can be designed for supporting data-driven decisions.