Social media recommendation systems increasingly trap users in geographical echo chambers. This platform dynamic optimizes for engagement, amplifying Western-centric content while systematically suppressing voices from the Global South and immediate regional neighbors. While users possess high cognitive awareness of these echo chambers, they suffer from an "agency deficit," lacking the tools to diversify their feeds. Through a Participatory Action Research (PAR) methodology, this study empirically mapped digital exposure gaps. Consequently, this thesis utilizes speculative Service Design to introduce OpenInfo, a strategic research NGO acting as a multi-sided ecosystem. Operating across progressive service layers, OpenInfo equips passive scrollers into active curators, collaborates with regional creators to bypass algorithmic suppression, and leverages empirical data to advocate for a participatory algorithmic filtration systems.