The framework for the connected HMI with respect to the configuration and computation of personalization data is presented. The connectivity medium of the future car would be based on either an embedded internet connection through the mobile data services, Bluetooth based data connection based on the user smartphone, Wi-Fi based connection using a wireless network connection or an optimized hybrid approach based on the availability of the connectivity medium or the preference of the user. It is assumed that the car cannot remain directly-connected at all times and the computational requirements of the advanced personalization application cannot be optimized using the inbuilt HMI hardware in the vehicle alone. Our algorithm is based on the concept of Constant Data Availability (CODA) distributed file system and the Distributed Application Processing framework (DAPF). The CODA distributed network filesystem is used for caching of data to perform disconnected operations during the non-availability of a suitable connectivity medium. The resources for the computational intensive personalization applications are realized through the DAPF. Both the offloading algorithm and the cloud model employed by the DAPF are described to detail the taxonomy of the framework while also analyzing its implications and critical aspects. Finally, the used cases for both personalized information and personalized entertainment are presented, which show the efficiency and usability of our framework for infotainment based applications. The configuration presented is scalable and has the capability to adapt according to network bandwidth.