The use of public open data is an opportunity for citizens and businesses to create products that contribute to decision-making, monitor and control public institutions as well as to improve the quality of life. Open data allows citizens to have unrestricted access to information collected by the government. Model-Driven Engineering (MDE) is an approach to software development that represent artifacts as models with the goal of reducing costs in the software development process. The use of these models allows to improve the processes of application development. The purpose of MDE is to try to reduce costs and development times and improve the quality of the systems, regardless the platform and guaranteeing business investments against the rapid evolution of technology. Thus, it will not be necessary to start from scratch whenever a new project is proposed or some type of maintenance on the product is desired so the associated cost will be reduced. This paper presents a metamodel and its corresponding domain specific language that captures public data and then transport, transform, analyze them in order to help decision-making to the agroindustry stakeholders.