Grant: Idea to Innovation

DaaSPatcher is a technology that allows data providers to segment and fine-grain control their Big Data and analytics users to run MapReduce (MR) jobs on some portion of the data provider 's Big Data. Brokers can act as intermediaries in the interaction, offering their own value-added transformations or subsets of the provider's data before it is processed by the user's MapReduce job. In this way DaasPatcher creates an ecosystem which is realized as a marketplace for Big Data sharing on the cloud. The current state of the DaaSPatcher requires advanced technical skills in order to segment the data, define the policies by writing specialized Maps, exporting data schemas and interfaces and finally writing the end user Map Reduce programs. In order to enable DaaSPatcher adoption we have to make the programming easier by creating a Software Developer Kit (SDK) and to transform the runtime platform into a more efficient release candidate. This proposal focuses on improving the performance, on in-memory processing and development tool support. Thus, a prototype can be demonstrated to a select group of customers. The feedback from these customers will enable the creation of a business strategy for a new Canadian-based, startup company in the rich media asset management industry. This startup will create new jobs, strengthen ties between industry and university labs, provide training for students.