QIRANA demonstration: real time scalable query pricing

AbstractThe last decade has seen a deluge in data collection and dissemination across a broad range of areas. This phenomena has led to creation of online data markets where entities engage in sale and purchase of data. In this scenario, the key challenge for the data market platform is to ensure that it allows real time, scalable, arbitrage-free pricing of user queries. At the same time, the platform needs to flexible enough for sellers in order to customize the setup of the data to be sold. In this paper, we describe the demonstration of Qirana, a light weight framework that implements query-based pricing at scale. The framework acts as a layer between the end users (buyers and sellers) and the database. Qirana's demonstration features that we highlight are: (i) allows sellers to choose from a variety of pricing functions based on their requirements and incorporates price points as a guide for query pricing; (ii) helps the seller set parameters by mocking workloads; (iii) buyers engage with the platform by directly asking queries and track their budget per dataset;. We demonstrate the tunable parameters of our framework over a real-world dataset, illustrating the promise of our approach.