<p>Tsunamis result in huge losses of lives and infrastructure located in coastal zones. Although tsunamis are rare, it is imperative to be prepared to cope with the adverse impacts of this hazard. The current study site, Nagapattinam district, was among the most severely affected regions in India during the 2004 Indian Ocean tsunami. Hence, this scenario has been considered to develop an efficient evacuation plan for the coastal communities in the selected region. The study presents a scalable and data-driven approach to objectively select suitable evacuation routes and estimate the walk time (evacuation time) required to reach safe shelters for people living in inundation zones. We obtained the modeled tsunami inundation data and spatial datasets, including building footprints, roads, and a high-resolution digital elevation model (DEM) derived from aerial data, as well as the road network from OpenStreetMap. The nearest school buildings outside the inundation zones are are automatically selected as safe shelters. Walk time from each building to the closest safe shelter was computed through a network-based routing analysis using roads and DEM is used to estimate walk time by accounting for slope factors. The buildings were categorized into four categories based on the walk-time and varying timeline of evacuation across the city was observed. The tsunami walk time maps produced here are vital for prioritizing tsunami evacuation planning and mitigation at micro-level to overcome bottlenecks/hurdles on existing infrastructure.</p>