Session: Project Alvarium – the Future of Big Data on the Edge

Big data has big trust problems on the edge.

The high level of trust assigned to enterprise data does not always extend to data born at the edge. The principles of enterprise trust insertion, however, can serve as guideposts for increasing edge data trustworthiness. This presentation introduces the motivation behind The LINUX Foundation’s Project Alvarium, as well as the building of the first Data Confidence Fabric (DCF). A DCF is a framework in which edge data is annotated with trust metrics as it flows towards applications. This talk will reveal the motivation behind Project Alvarium, describe the trust measurement techniques used within a DCF, and highlight how the value of the resulting data increases.

Project Alvarium will focus on … “building the concept of a Data Confidence Fabric (DCF) to facilitate measurable trust and confidence in data and applications spanning heterogeneous systems. Alvarium is a Latin word for “beehive”, in which the community works in an environment of trust to accomplish a common goal: build an open Data Confidence Fabric.

When enterprise data flows between storage systems and applications, it is handled by various layers of hardware and software trust insertion as part of a data path or stack. This stack is often created and maintained by highly-trained enterprise architects and security practitioners. As such the data is implicitly trusted when it arrives at an application.

Extending this implicit trust to edge and IoT data is challenging. The data flows over a wide variety of heterogeneous technology and networks, across an ecosystem so large that it is often beyond the reach and capabilities of an enterprise security team.

Applications, therefore, must learn to trust edge and IoT data explicitly by examining trusted annotations that accompany the data. These “trust insertions” generate metadata and scores that are forwarded alongside the data. Annotations, when they arrive at an application, allow the confidence of that data to be “measured.”

Higher confidence scores lead to higher value: risk is reduced (e.g., regulatory fines) and new forms of revenue from data become possible (e.g., data marketplaces).

The presentation will describe the results achieved from building and running the industry’s first Data Confidence Fabric and will include an invitation to join the Alvarium open community and work together on trust configuration, annotation, and scoring methodologies.