Earth Observation missions are entering a new era characterized by growing demand for near real-time, reliable and high-quality data. In this context, satellites agility, short delivery times, and efficient use of existing ground and relay infrastructure have become key performance drivers. This thesis addresses these challenges by developing and applying a modular simulation framework capable of modeling agile satellite constellations from end to end - from orbital propagation and target access to task scheduling, communication events, and latency computation. At its core is a Python-based tool designed to assess end-to-end latency and compare Earth Observation constellations under realistic operational conditions. The work was carried out at OHB System AG, one of Europe’s leading space prime contractors, within its Pre-development Department, which focuses on early-stage feasibility studies and mission proposals.