As protests spread across Khartoum during the 2019 revolution, civilians gathered in a mass sit-in demanding the end of a 30-year dictatorship. The movement was met with escalating violence, culminating in targeted attacks on protesters, including the violent dispersal of the sit-in on June 3rd. Violence unfolded in real time, yet there were no mechanisms capable of documenting, verifying, or responding to it at the speed at which it occurred.
At that time, Maysaa Sati was an architecture student. In response to this escalation of violence, she started working on human rights documentation as an early open-source intelligence researcher. Her work involved verifying fragmented digital signals, geolocating photographs, and corroborating testimonies to produce evidence for international accountability initiatives. This included preparing rigorous documentation intended for the International Criminal Court. She was taking chaotic, ground-level testimony and turning it into structured, verifiable evidence.
That experience shifted her professional trajectory. Working directly with affected communities in Khartoum exposed the limits of existing systems in capturing and responding to violence as it unfolds. She spent months studying displacement camps in Darfur for her architectural thesis, analyzing how spatial design could improve living conditions for long-term displaced populations. She wanted to understand what pushed people to move and why systems designed to protect them consistently failed to do so.
When she later arrived at the Massachusetts Institute of Technology, her perspective evolved further. The fundamental problem facing vulnerable populations was not strictly spatial but systemic. In some cases, critical data simply does not exist. In others, it exists in abundance but remains fragmented, inaccessible, or unused. The failure lies across both stages: how data is collected, and how it is translated into action.
“Across conflict, displacement, and climate disasters, there is always a gap,” Sati explains. “Sometimes the data isn’t there. But even when it is, it’s rarely structured or delivered in ways that lead to timely, meaningful action for the people it is meant to protect.”
This insight led her to found AcasiaPulse through the MIT DesignX accelerator program. The project used artificial intelligence and social media to predict displacement patterns before they showed up in official reporting. Typically, humanitarian organizations rely on delayed demographic statistics and official border crossings to allocate resources. AcasiaPulse took a different approach. Machine-learning models looked at digital signals, keywords, and online communication patterns to reduce the time between early crisis signs and the actual humanitarian response. Sati was building systems to predict human movement rather than simply reacting after populations had already been displaced and traumatized.
Today, Sati applies this same thinking to climate resilience. She serves as Co-Founder and Chief AI Officer of FloodLine, a startup building a mobile-first flood preparedness platform for small businesses. While the platform provides localized guidance during heavy rainfall, her focus is on the underlying data systems. Her role centers on ensuring that fragmented, hyperlocal information can be accessed, verified, and used in real time.
At the core of her work is the development of a retrieval-based system that allows small businesses to interact with a chatbot grounded in verified sources. Rather than relying on general-purpose tools, the system draws from local reporting streams and federal sources like Federal Emergency Management Agency. The goal is to provide information that is not only relevant but specific to location, context, and moment of need.
This work builds on an earlier version of the platform developed by FloodLine’s founding team. Sati’s role has been to extend that foundation, structuring how information is retrieved, validated, and delivered so that it can function reliably during emergencies. The emphasis is not on generating new information, but on ensuring that existing knowledge can be accessed in ways that are immediate and actionable.
In parallel, she is contributing to the longer-term development of FloodLine’s data infrastructure. As more user-generated and institutional data is collected, the aim is to build structured datasets that can support more predictive and analytical capabilities over time. These systems are being designed to inform not only individual users, but also planners and public agencies working on flood risk and preparedness.
Technology companies frequently assume that larger datasets will automatically yield better humanitarian outcomes. Sati argues that the issue is not simply the availability of data, but how it is used. In some cases, critical data is missing. In others, it exists in abundance but remains fragmented across agencies or inaccessible to those who need it most. Even where predictive models are possible, their outputs are rarely translated into systems that support timely, ground-level action.
At the MIT Leventhal Center for Advanced Urbanism, Sati continues to explore how qualitative and spatial data can be structured into usable intelligence. Her work consistently bridges the gap between lived human experience and computational modeling. She focuses on building systems that translate the chaos of a crisis into organized, ready-to-use information for policymakers and emergency responders. By applying architectural principles to data architecture, she creates frameworks that respect the physical realities of the built environment while leveraging the speed and scale of artificial intelligence.
Her work shows that building AI for crisis response requires changing how information is collected and delivered at the ground level. Resilience fails when information fails. Sati is building the infrastructure to prevent that failure, ensuring that the data generated by communities can be used to protect them.
FloodLine co-founder and Chief AI Officer, Maysaa Sati
