Every day, thousands of women board Guatemala City's Transmetro aware of what risks they are exposing themselves to. This thesis explores why sexual harassment persists as a structural condition of that system despite the existence of an institutional response, and proposes three interconnected service design interventions aimed at interrupting the conditions that sustain it.The research adopts a pragmatist orientation grounded in Systems Oriented Design, combining secondary data analysis, semi-structured interviews and conversations with women users and Transmetro workers, workers from Municipalidad de Guatemala, ride-along observation, a service safeguarding audit, and a participatory design workshop. The central fi nding of this analysis is not the absence of institutional effort because Municipalidad de Guatemala produced a protocol addressing gender-based violence in public transport in 2023, but the structural invisibility of that effort and its consistent failure to reach the women it is meant to protect.Through cross-method synthesis, the research identifi es two interlocked reinforcing loops sustaining harassment as a systemic condition: a structural failure loop, in which inaccessible reporting mechanisms prevent data generation and therefore prevent institutional response; and a withdrawal and invisibility loop, in which women's behavioural adaptations are misread by the system as evidence of safety rather than avoidance. Both loops converge on the same point — women's silence — which the system consistently interprets as reassurance. The thesis names this dynamic performative safety: the misalignment between what an institution claims to offer and what women experience in practice.2In response to these fi ndings, three design proposals are developed. The fi rst, a Protocol Accountability System, makes the Municipalidad's existing commitments publicly visible and verifi able through a partnership with OCACGT, the most trusted civil society actor on this issue in Guatemala. The second, a Gender-Responsive Staffi ng Framework, addresses the human infrastructure through which any protocol is either delivered or not, positioning female staff at the highest-risk touchpoints identifi ed through the research. The third, a Low-Threshold Experience Channel, creates a non-punitive pathway through which women can share experiences that fall below the threshold of formal complaint, generating population-level evidence without requiring institutional trust they have not yet been given reason to extend.Together, these proposals do not claim to solve the problem of harassment in the Transmetro system. They propose a set of structural changes that make the system less capable of sustaining harm invisibly, and more accountable to the commitments it has already made.