Today’s users want to access their data everywhere and any time – in various environments and occasions. The data itself can be very complex – the problem is then in providing such a complex data to the user with some interaction limitations implied by current working environment (for example noise preventing user from receiving sound information), abilities of end device (PDA unable to display large images), specific needs of user group (e.g. visually impaired user) etc. One of the possible solutions can be in allowing interaction and providing data in different modalities – audio (speech or act” sounds), visual, tactile etc. Such a multimodal system must also cope with the dynamically changing conditions (e.g. working environment – generally with the user context). Our project is focused on definition of an architecture allowing users to access complex data using different input/output methods such as images from digital camera, speech or haptic feedback as output from t...