We analyzed the information system available for the search, identification and delivery of human remains in the Colombian conflict, based on the social studies on science and technology, and on ethnographic material. We understand that in this system the convergence of memory practices with the power to reconfigure past (and present) by means of selection and exclusion processes framed in scientific practices with character of truth is present. We reconstruct the path for the production of information and identify the filters that work when classifying and analyzing it under scientific standards. We follow the trails that the excluded information leaves and that may be potentially rescued. We question ourselves about the memory built from those filters and traces. We argument that it is vital striving to have the system open the way for uncertainty, so it contributes to the production of a collective memory of the conflict that is multiple and thorough.