FIMS has demonstrated how to run a workflow on Amazon at NAB 2017. Thius presentation explains the different parts of the implementation using S3 (storage of content and associated metadata), step functions, API Gateway (REST), lambda functions and dynamoDB for temporary storage of metadata. The workflow at NAB consisted of validating semantic data associated with the editorial object and hires essence, the automatic extraction of technical metadata using mediainfo, the creation of a lowres proxy, the extraction of a thumbnail and the push of the aggregated metadata into a semantic repository at the end of a series of processes orchestrated by step functions. This presentation reported on lessons learned and future activities in FIMS.
Presenter(s): Jean-Pierre Evain (EBU)
Event: MDN Workshop 2017
