A Practical Response to the MovieLabs 2030 Vision

Margaret Craig, CEO of Signiant, Ian Hamilton, CTO of Signiant

Executive Summary

The Signiant Platform touches most of the world’s professionally produced media content at some point during its life cycle, providing mission-critical functionality that extends far beyond conventional file transfer. The ambitious future state outlined in the MovieLabs 2030 Vision whitepaper, titled The Evolution of Media Creation, generally aligns with the Signiant roadmap — and it is not just a theoretical vision for us. Halfway through the proposed 10-year journey, we can provide insight into a fully functional and widely deployed software implementation that embodies many aspects of the framework provided by MovieLabs.   

This short paper outlines specific real-world considerations relating to Section 3.1 of the 2030 whitepaper, which is called A New Cloud Foundation. Signiant also has direct involvement in the security and workflow facets of content creation, but we are containing the scope of this discussion to the first three MovieLabs principles. These relate to the active (non-archive) aspects of technical infrastructure and are stated as follows: 

1. All assets are created or ingested straight into the cloud and do not need to be moved.  

2. Applications come to the media. 

3. Propagation and distribution of assets is a “publish” function.  

On the following pages, we put forward twelve practical guidelines for next-generation media technology infrastructure that relate directly to the above 2030 principles and are aimed broadly at anyone in the industry who is involved with forward-looking architectures and technical strategy. The guidelines are grouped into three categories and presented in a logical flow, but each point can nominally stand on its own.   

File Storage & Data Movement 

1. All media assets must securely reside in Internet-accessible storage.   

2. The next-generation tech stack must fully support storage diversity.   

3. Users and applications require federated access to multiple asset repositories. 

4. Fast, reliable, secure data movement is an enduring foundational service.   

5. All file movement is not created equal.     

6. On-demand proxies can provide efficient access with enhanced security.  

The Evolution of Applications (through an infrastructure lens)  

7. Ingest into proprietary systems is generally unnecessary.   

8. Browser-based tools with adaptive backends will replace the virtual workstation model.  

9. Edge-enabled semantic search will trump cloud AI/ML services that enrich standard metadata.  

10. Moving the automated workflow model from code to configuration will enable more efficient media fulfillment.   

Technology Delivery  

11. Only multi-tenant SaaS can deliver scalable, resilient functionality at the lowest possible TCO.  
12. An industry-wide connective fabric for M&E already exists in the form of the Signiant Platform. 

Context & Signiant Perspective 

As active participants in the media ecosystem, we were eager to digest the MovieLabs 2030 Vision whitepaper when it was first published in 2019 — and studio customers were immediately curious about our reaction.  Signiant moves files, so surely the company would be at odds with a view of the future that leads with “files do not need to be moved”? Actually, no. We see huge resonance between the 2030 Vision, the needs of the industry, and the capabilities of the evolving Signiant Platform. While we do think the focus on eliminating file movement is something of a red herring, most of the concepts align with our perspective on the future.   

The hard part, of course, is real-world implementation of these principles. This is where innovative software companies like Signiant come into the picture. As a pioneer in the application of cloud technology within the media sector, we have hands-on experience (and lots of data) relating to the ways in which media companies really employ the cloud today and how adoption is trending. We’re intimately familiar with the elusiveness of standards in the media world, and we see firsthand how legacy workflow automation is collapsing under its own weight. Our platform is in the thick of the metadata revolution, where classic structured data schemas can now be augmented or even replaced by unstructured data and AI/ML services. It is a constant challenge to align product development efforts with market forces, but the rapid growth of our business since the advent of cloud and SaaS is a testament to the real value that customers derive from our company’s targeted innovations. 

In other words, we speak with some authority from the technology trenches of the industry — and Signiant has a track record of operating at the leading edge. These notes are aimed at our executive peers in both the tech supplier and M&E customer communities; they are not intended to be broadly accessible outside the sphere of media technology. Unencumbered by the MovieLabs imperative to present an agnostic and generalized outline, we’re intentionally providing a Signiant-specific point of view and tying our commentary to real initiatives within our company. Through this focused lens, we hope to contribute to the dialogue that will make the 2030 Vision a reality. 

Twelve Real-World Guidelines

File Storage & Data Movement 

1. To realize a next-generation approach to content creation, all media assets must securely reside in Internet-accessible storage. This is simply a more pragmatic re-statement of Principle 1 in the whitepaper, and it is consistent with the paper’s definition of ‘cloud’ at the beginning of Section 3.1. By removing that word altogether, we avoid conflating the business models and specific services of AWS, Microsoft Azure, and Google Cloud Platform with an overall vision for essence and metadata flow across distributed infrastructure. Clearly, many customers will eventually choose to utilize one or more of the hyperscale public cloud platforms, but that is a separate concern and should be regarded as such.   
   • Signiant plays a key role in this aspect of the 2030 Vision by providing a common internet-addressable interface to all storage, regardless of whether it is natively connected. This core capability of the platform is enabled by customer-installed software in on-premises data centers and Signiant-operated software in multiple public cloud data centers, which together form a global mesh of content repositories that can be accessed via the Web.   

2. In alignment with the multi-cloud, hybrid cloud storage strategies employed by most media enterprises, the next-generation tech stack must fully support storage diversity. This is particularly true when considering the entire content lifecycle from creation to archive. The MovieLabs whitepaper glosses over the storage independence dimension, but the industry would view it as untenable to optimize around the technical specs or economics of a specific public cloud platform.   
   • The Signiant Platform supports all storage types, and customers consistently reinforce how critical this built-in optionality is to their architectures. Taken at face value, the 2030 whitepaper seems to promote a high level of ‘data gravity’, which refers to concentrating much of the data in one place. This might actually be problematic, since conventional wisdom holds that data gravity tends to increase cost and complexity — and distributed storage is often implemented to mitigate those challenges. Our software enables storage architectures that are logically centralized rather than physically centralized, thereby taking data gravity concerns out of the equation. Content can be distributed amongst multiple storage types and locations, based on whatever criteria matter most in each situation, and still present as a single storage system to applications, users, and workflows. 
3. Users and applications require federated access to multiple asset repositories with diverse storage types and media formats. This enables the physical location of media to be abstracted for users and accommodates the reality that media assets are rarely in a single storage location in a single format. Even in a straight-to-the-cloud future, different assets may go straight to different cloud locations. Applications must be able to operate in an environment with distributed assets as well as distributed users, something they can do natively or by relying on a third-party service. It is important to note that in order to be useful, federated access must be essentially transparent to end users, which in turn means that it must be fast, reliable and secure.
   • Signiant’s foundational transport technology enables our platform to serve as the storage abstraction layer for distributed media asset environments. Our network optimization software, along with associated compute resource, is deployed at every storage repository. Users and applications are thereby assured that all access operations will perform as if local, regardless of storage type or location. (This same technology, of course, underpins conventional ‘move file’ operations.) We’re now adding applications to the platform that provide cross-storage capabilities directly to the end user, for example federated search in our Media Engine offering.
4. Fast, reliable, secure data movement is an enduring foundational service that will underpin storage diversity and application choice in the long term, and facilitate efficient ingest, collaboration, and distribution in the short/medium term. By stating that “files do not need to be moved”, we trust that MovieLabs is emphasizing the need to minimize unnecessary file movement rather than literally proposing that elimination of all file movement is an appropriate existential goal for the industry. There are far more significant cost and efficiency drivers in our content creation processes, and ‘all files sit still’ is unachievable in practice. At the most fundamental level, raw footage is not captured in the cloud and the creative professionals who operate on files aren’t in the cloud — so clearly these two realities need to be accounted for. Camera-to-cloud workflows move a huge amount of data, and there is a very fine line between streaming from the camera or sending files from the set. We definitely agree that the file movement landscape will continue to evolve. It will become less necessary to move files simply to provide another party with access. Modern applications will increasingly abstract file movement away from the user, so over time there may be a shift away from user-centric file transfer tools and toward API-driven data movement. And data transfer can certainly be minimized for captive workflows within a single hyperscale public cloud platform. 
   • With the Signiant Platform, people and systems can move files of any size, anywhere in the world, over any IP network. This is made possible by our network optimization technology, which minimizes the impact of latency and packet loss to ensure that all available bandwidth is fully utilized — and the same technology facilitates the movement of both files and streams. The laws of physics can’t just be ignored. At one point in the MovieLabs whitepaper, the reader is advised to “envision a high bandwidth, low latency production environment”. Sure, but with global production you don’t have control over latency. The speed of light is the speed of light, hence the need for Signiant’s network optimization technology. 
   • Data from the Signiant Platform bears testament to the continuing criticality of file transfer in the cloud era. We saw a 3.6-fold increase in file movement in and out of AWS/Azure/GCP from 2020 through 2024, with roughly equal amounts being uploaded and downloaded.
5. Having established that files must sometimes move, it is important to also review the corollary that all file movement is not created equal. Rather than stating that assets “do not need to be moved”, it is more useful to characterize the goal as minimizing copies of files — but there are nuances there too. Making copies of files solely for access purposes is bad. The MovieLabs paper notes the problems of rogue copies and version control, and extra copies also drive cost and sustainability problems. Keeping a grip on that one authoritative version is critical, but it is likewise essential to be able to move media or create ephemeral derivative copies for purposes of efficiency and performance. In this context, focus shifts to the various ways in which efficiency of access can be optimized. One form is optimization of network protocols for bit-accurate delivery of both streams and files. Another is optimization of a user-specific stream by tuning the format for a specific task. Yet another way to optimize access is to cache ephemeral copies of the relevant content in a transparent fashion. Both moving and copying can be good or bad, depending on the situation — so both capabilities must be available and both must be carefully managed in concert with applications.    
   • Our transport protocol is very powerful, but it isn’t a standalone technology. The Signiant Platform delivers customer value by pairing the underlying infrastructure software with an increasingly sophisticated and adaptive application stack. We often describe our SaaS architecture as being ‘SDN-like’, in that it features a cloud control plane paired with a storage-connected data plane. This analogy works as far as it goes, but for advanced M&E use cases it is instructive to think of the control plane and data plane as being overlaid by a ‘media intelligence framework’ that unifies the entire system in a media-aware fashion. This allows us to do things like correlate metadata and essence, dynamically align infrastructure with use cases, and situationally address all the various forms of media access optimization outlined above. The Signiant Platform is uniquely suited to enable the intelligent, media-aware move/copy operations that will underpin the future of content creation.
6. Since the advent of file-based media, efficient interaction with video essence has been achieved by providing users with lower-resolution proxies that are pre-generated and stored in an accessible location. This is a workable solution, but it does raise concerns about security, storage costs, and copy proliferation — and making those proxies accessible to distributed users can be a challenge. There are emerging opportunities for improvement by looking holistically at media access for both heavy assets and proxies in a distributed production environment. With modern technology, a system that has access to the heavy asset can generate a proxy file on demand and deliver it as a unique ephemeral stream to each user. These on-demand proxies can provide efficient access with enhanced security and offer various other benefits.  Proxy streams can be customized for both security requirements and use case — for example, frame-accurate clipping demands much higher resolution than simple content verification. It is straightforward to add user-specific visual spoiling or watermarking to individual streams.   
   • The Signiant Platform now includes on-demand proxies as part of the Media Engine offering, enabled by our existing software footprint at content repositories. The allows browser-based user interaction (play, clip, send) with media assets that remain stored in their original format. With broad format support including camera RAW, customers are gravitating to this basic tool for use cases ranging from simple dailies viewing to interactive media storage in S3 when full creative collaboration toolsets are overkill. Having established this baseline of generating a unique ephemeral per-user stream, it will be a natural extension to build out more adaptive capabilities for various use cases and security requirements. 

The Evolution of Applications (through an infrastructure lens)  

7. Once all assets are in Internet-accessible storage, labor-intensive ingest into proprietary systems is generally unnecessary. Assets can remain in their original storage locations, with users and applications accessing them as necessary. Media asset management functionality is evolving in parallel, with modern approaches to metadata capture and search further diminishing the need for ingestion into legacy heavyweight MAM systems. A more nuanced view of “applications come to the media” can help make all of this implementable in practice. Arguably, not all application elements are created equal and some have a greater need than others for proximity to the media. If software functionality that requires access to an entire repository is close to the storage, there is more flexibility in terms of where the creative application itself resides. 
   • The Signiant Platform inherently “comes to the media” — our software and the associated compute resources are already deployed at most content repositories to facilitate file access and transfer. Via this edge footprint we can extract metadata, index the storage, and produce on-demand proxies for the powerful search and preview service called Media Engine. Creative professionals working with on-premises applications can find the content of interest and selectively bring a clip or a pre-vetted heavy asset from a remote cloud repository down to their desktop. There’s no need to move or ingest the entire source library, and full cloud-based editing isn’t always needed. 
8. In another example of how applications and infrastructure must evolve together, we believe that browser-based tools with adaptive backends will replace the virtual workstation model. The notion of remoting screens is a workable interim approach for retrofitting desktop tools to an Internet-based world, but cutting-edge approaches will leverage the evolving infrastructure discussed in this paper. Browser-based tools can be designed with sophisticated backends that are aware of the task being performed and take action accordingly. For example, it may make sense to cache non-authoritative copies of work in progress for efficient access. This can be done in a manner that is completely transparent to the user, with the ephemeral copies being created, cached, and eventually deleted without any user involvement.   
   • The Signiant Platform is a behind-the-scenes operator in this scenario. Via API, a creative application’s backend can use our technology to transparently access, move, copy, and delete files located anywhere in the world, in any kind of storage. The creative professional’s user experience will be optimized in a very adaptive manner, ensuring that their work is unencumbered by infrastructure concerns. As the bright line between applications and infrastructure blurs, we anticipate working ever more closely with application providers to create these optimized creative experiences.   
9. As noted in # 7, ‘edge services’ can be a powerful tool in a reimagined media universe. (n.b. We are using ‘edge’ in a B-to-B storage-facing sense, not the consumer-facing CDN cache sense that appears at some points in the MovieLabs whitepaper.) In some ways it’s a harder problem to find assets if they might be in multiple locations, but the good news is that some of the heaving lifting can occur close to each content repository, on a distributed basis. We see a very direct path for leveraging our existing edge footprint in the practical application of new AI/ML capabilities and think it can be argued that edge-enabled semantic search will trump cloud AI/ML services that enrich standard metadata. While standard metadata enrichment has the advantage of being backward compatible with existing search mechanisms, it can’t capture the essence of what’s represented in an image like an ML-generated semantic vector embedding can. This powerful new approach will also feature lower costs and there will be far less friction in deployment.   
   • Plans for Media Engine include adding AI-powered semantic video search, which will allow users to find content by searching on descriptive phrases without the imperative for specific words to be present in metadata. Our working prototype leverages the CLIP model from OpenAI, and the results are mind blowing. We’re leveraging three system elements that are already part of the Signiant Platform:  
     • Media Engine’s existing video search framework and UX 
     • Software and compute resources already deployed at content repositories, providing infrastructure necessary to generate CLIP embeddings  
     • Thumbnail extraction functionality that can be leveraged for the embedding step
10. MovieLabs posits that distribution efficiencies will eventually be possible via a ‘permissioned access’ or ‘publish’ model, and the Signiant Platform is moving in this direction. Our secure inter-company permissions framework allows customers to share access to an asset only for a specific task (download, play, clip, etc.), which lays the groundwork for a more comprehensive publishing model. In the near term, there is some low-hanging fruit that aligns with the software-defined workflow aspects of the 2030 paper. Specifically, we believe that moving the automated workflow model from code to configuration will enable more efficient media fulfillment. With each content distributor tightly prescribing exactly how essence and metadata must be delivered to them, there’s no good reason for every contributing media entity to create their own bespoke system for content preparation. Forward-thinking media organizations will let go of highly customized workflows and lean into an off-the-shelf mindset for their fulfillment operations, thereby driving cost and efficiency gains.   
    • Signiant is building next-generation process automation into our existing SaaS platform, which due to its multi-tenant nature can deliver the scalability and resiliency that are required for at-scale multi-platform distribution. Our unique position in the media technology stack directly enables this expansion: we have access to content repositories, the mechanism to deliver assets to their final destinations, and existing operational touchpoints with both essence and metadata. In 2021 we acquired Reach Engine by Levels Beyond, a MAM and workflow orchestration offering that provides high-volume, high-velocity content processing for some of the world’s largest media companies. Through this acquisition we gained talent, technology, and an existing customer base that paved the way for a next-generation Signiant initiative code-named Media Factory.   
    • At the core of Media Factory is a productized multi-tenant content fulfillment pipeline that progresses the essence and metadata through a series of standard stages. This is not an individual hardcoded workflow, but rather a series of dynamically reconfigurable modules that deliver core functionality in a scalable and resilient manner. For each individual content operation, the modules are dynamically directed by a policy engine that understands available inputs (via a customer-delivered workorder) and required outputs (via a codified delivery spec). In other words, these are software-defined workflows — but Signiant does a lot of the heavy lifting so that customers have no need to continually reinvent the wheel.    

Technology Delivery

11. The MovieLabs whitepaper repeatedly cites the cloud as a key enabler of the future state, but it neglects to note the power of multi-tenant SaaS — one of the most fundamental ways in which customer value is delivered in a cloud-centric world. Direct consumption of IaaS (compute and storage) is straightforward, but multi-tenant SaaS is a different proposition. It is very complex for suppliers to build and operate, but very easy for customers to consume. Within the broader technology world, the fact that only multi-tenant SaaS can deliver scalable, resilient functionality at the lowest possible TCO is accepted and understood. It’s time for the media industry to follow suit. Lift-&-shift deployments and single-tenant cloud systems are sometimes referred to as ‘cloud-native’, but they are not true SaaS and do not deliver SaaS benefits. Multi-tenancy is a critical consideration for both creative applications and infrastructure in the next-generation media tech stack.  
    • Signiant has led the way in multi-tenant SaaS within the media tech sector. Our patented architecture includes a multi-tenant control plane as well as multi-tenant data plane endpoints in every AWS, Azure, and GCP region. The technical elements of scalability and resiliency are just part of why the industry has gravitated to this platform. Once customers experience a system that is super easy to deploy and manage — and delivers fast innovation and fast time to value — there is no going back. SaaS is the most accessible form of cloud and it’s here to stay.
12. In addition to the technical and operational benefits noted above, multi-tenant SaaS intrinsically provides a means of connecting and enabling the natural ecosystem that underpins a vertical industry like Media & Entertainment. The latter parts of the MovieLabs whitepaper focus on the need for standards to enable security, access, and tracking of media elements across the industry. We understand the proposition and will certainly support standards if they emerge, but if history is any guide it will not be straightforward to achieve broad adoption. In the meantime, an industry-wide connective fabric for M&E already exists in the form of the Signiant Platform.  It is designed and operated by a trusted, ubiquitous incumbent supplier that is very intentional about serving companies of all sizes.      
    • The Signiant Platform is the mechanism of choice for content exchange within and between media companies, something we can achieve either via file movement or via secure access permissions. This was possible pre-SaaS, but the multi-tenant control plane now makes inter-company handshakes much easier and more secure. We can envision dozens of other ways in which this existing connective tissue can be leveraged to deliver on various elements of the 2030 Vision.   
    • Is it problematic for one company to play such a central role? We’re mindful of this concern and therefore make sure we play by the rules of true platform companies. Signiant is delivering on a platform strategy in the purest sense: we serve as the core of an ecosystem, participants benefit from the presence of others, and our system is operating at scale. True platforms inherently need to remain open — think Salesforce.com. Customers can adopt our platform secure in the knowledge that we aren’t trying to boil the ocean or control everything. We sell optional modules rather than end-to-end systems, we position our platform as a conduit for metadata rather than a single source of truth, and customer storage is always a natural hand-off point for other systems. 

Summary

As the MovieLabs 2030 Vision comes into focus, this paper offers food for thought from the vantage point of the Signiant Platform, a multi-tenant SaaS offering that is deeply embedded in today’s global content creation workflows and ideally suited to help the industry move into the future. Key takeaways:  

• Here’s a recap of how to think about Signiant’s role in the modern media technology stack. Our software needs access to storage in order to send large media files. Once this storage access footprint is established, there are lots of other value-add things that the software is uniquely situated to do — relating to essence, metadata, and the linkages between the two. We are steadily building adjacent media-centric SaaS services and delivering them in a very low-friction way to our 1000+ M&E customers. 
• We believe that infrastructure and applications need to come together in increasingly sophisticated, media-aware ways to optimize next-generation content creation. We’re leaning into partner conversations on this front and welcome engagement with anyone in the industry with an interest in this evolving dimension. 
• The devil is always in the details. There are critical subtleties in accessing/transferring/copying/streaming files. It’s therefore important to be precise with cloud and SaaS terminology, and there are real practical constraints around every corner. In specific areas where we’re the experts, we are always happy to compare notes and share lessons learned.   
• There are exciting opportunities for innovation in everything from proxy delivery to high-volume content fulfillment to semantic search. Signiant is always pushing the envelope to leverage our technology and market footprint to deliver real differentiated value to our customers.    
• Finally, we welcome industry response and debate regarding the ideas and opinions we have laid out in this paper. We admire MovieLabs for putting a stake in the ground with the 2030 Vision and facilitating dialogue on related topics, and it is in that spirit that we wrote this paper.