Our goal was to design a modular, scalable and cross-platform system providing all necessary features to support image-guided interventions for neurosurgery and radiosurgery activity: patient and center management, clinical knowledge system, archiving, report production. We used a server-side web application based on open, standard and robust technologies: Java, XML, JDO and persistence layer (Castor), SQL-transaction enabled Database, HL7, DICOM3.0 and Apache projects. Object orientation and design patterns enabled efficient development strategies: Model-View-Controller approach, document generation engine (PDF), secure authentication model. Two years ago, we launched a simple agenda application and started to deploy progressively additional components. Now the actual framework is used by more than 30 users to encode treatment data (654 interventions) and patient follow-up (597) for diverse image-guided procedures (radiosurgery, surgery navigation, intraoperative MRI). Extracted statistics were presented in more than twelve scientific communications. Daily administrative documents are automatically produced. Planning and treatment data including images are collected from various medical workstations and archived both online and offline with fully automated management.