The FlowEdge uses local and less expensive on-premise storage for content that does not need to be shared with many other users or viewers, and instead places content into its cloud module for bigger audiences and high volume accessibility, usually automatically via the internal project and distribution mechanisms, but also manually for quick access content. It also manages the online content automatically based on access numbers, user interactions, playout times or any other logical rulesets.

The FlowEdge makes use of cloud servers for the online modules, but also the local processing power, including a GPU for transcoding and AI services for serving the internal network. The dedicated GPU can transcode videos in FullHD with up to 500 frames per second, much faster than any cloud transcoding service, before transferring it to other destinations. All FlowEdge instances act as a single system with a single sign on, while using every processor and dedicated GPU processor available in the complete system, summing up all the computing power and network resources.

Content is processed very fast on the local GPU and then transferred to external cloud modules with a format and size optimized and downsized for online purposes or online delivery. Content that is not leaving the on-premise area is kept in its best quality and resolution for archiving and later re-use. Content that has to be analyzed by AI services is downsized massively to archive very fast transfer rates to external AI services, while still receiving the perfect analysis and metadata. Media that needs to be distributed to OTT, VOD or CDN platforms is transcoded and edited to the best fit quality and then distributed to its destination. At no time will the FlowEdge transfer huge original-sized camera or editing material to online services or distribution targets, thus saving enormous transfer volumes, making the overall system extremely fast by eliminating and bypassing bottlenecks.

The DNA of FlowEdge already has scalability built in it. This is what we refer to as superscalability, meaning that it has a layout of a decentralized media asset management system, being able to connect to one or more other instances via the Flow2Flow connector API and work as a single instance, sharing content and metadata across a globally distributed network (horizontal), while storing content for distribution, online collaboration and OTT/CDN/VOD viewers in cloud instances (vertical). This scales in both directions: horizontally and vertically. This multi-node setup also accumulates the processing power and network capabilities of many single computers to a very powerful single system, while still being able to drive completely independent cloud platforms within its own boundaries, extending the environment into the mentioned hybrid cloud/on-premise setup.

Confidential content or media that is still work in progress is always stored locally on-premise with higher security levels, not accessible from the outside world or on a cloud system. The content that is used for big audiences and larger numbers of viewers, including the media for online collaboration is stored in the cloud modules of the Flow.Edge, providing highest speeds and huge volume transfers for many parallel views.