The mining process is responsible for generating knowledge and processes to find from the records that are extracted in information systems. This paper develops this perspective, low integration methodology business process automation, CRIPS_DM methodology, focused on data mining and process redesign methodology BPTrends. With this combination methodologies is to identify factors, patterns, variables and bottlenecks that have great impact on cycle times that directly affect the performance the process control and national and international procurement goods and services for research projects of the Pontificia Universidad Javeriana.