Industry Insights: The next generation of workflows is intelligence-driven

As media workflows become increasingly software-driven, metadata, automation and operational intelligence are emerging as the foundation for how content is managed, distributed and monetized.

In the final installment of this three-part Industry Insights roundtable on workflows and integration, broadcast technology leaders examine how organizations are building more intelligent media supply chains designed to scale across formats, platforms and rapidly evolving distribution demands.

The discussion explores how metadata is being generated and preserved across production pipelines, why lifecycle management and version control are becoming more complex and how organizations are measuring workflow modernization beyond simple cost savings. Participants also examine the growing role of AI-driven orchestration, predictive automation and data-centric operations as broadcasters move toward workflows that are increasingly adaptive, context-aware and capable of responding dynamically to changing operational conditions.

Key takeaways from this Industry Insights roundtable

• Metadata expands: Metadata is increasingly generated automatically through AI-powered analysis, transcription and contextual enrichment throughout the production pipeline.
• Success evolves: Broadcasters are measuring modernization through efficiency, scalability, collaboration and operational flexibility rather than hardware reduction alone.
• Lifecycle complexity grows: Version control and content lifecycle management are becoming more difficult as assets expand across formats, regions, languages and distribution platforms.
• Automation matures: Workflow automation is evolving beyond predefined tasks toward more adaptive, intelligence-driven orchestration capable of responding dynamically to operational conditions.
• Data becomes central: The effectiveness of future AI-enabled workflows will depend heavily on the quality, accessibility and governance of metadata and operational data across the media supply chain.

How is metadata being generated, enriched and preserved as content moves through production and distribution pipelines?

Thomas Carlisle, senior solutions architect, TAG Video Systems: Metadata integrity is a QC problem as much as a data management problem. Errors or gaps introduced early in a pipeline, like wrong language tags or missing audio descriptors, compound as content moves downstream. Automated QC at multiple pipeline stages is the most reliable way to catch metadata drift before it affects delivery.

Guillaume Aubuchon, VP, product management, Avid: Metadata is king. It’s increasingly generated automatically through transcription, analysis, and contextual signals from editorial tools and then enriched as content progresses. The key shift is treating metadata as persistent and portable, not bound to a single system. Platforms are being designed to index, preserve, and propagate metadata across tools and environments so that editorial context isn’t lost as content moves between on‑prem, cloud, and partner ecosystems.

Derek Barrilleaux, CEO, Projective: For rushes, information like logged data and technical metadata can be taken from the camera and automatically transferred to the PAM for editing, where creative teams can use it to search. Project-based metadata also helps users find content based on project context. Recent innovations allow users to leverage semantic and visual search, as well as searchable transcripts, to augment traditional metadata. And, of course, it should be possible to share all this metadata both up- and downstream.

Geoff Stedman, CMO, SDVI: Metadata generation has evolved significantly — where it was once largely a manual entry task at ingest, AI-powered tools can now produce rich, time-based metadata, including speech recognition, scene classification, and compliance flagging, all without human intervention. The challenge we see most often is metadata preservation: content passes through multiple systems — MAM, transcoding, QC, distribution platforms — and there’s no guarantee that the metadata from one step survives the handoff to the next. Orchestration platforms that treat metadata as a first-class object — capturing it, transforming it as needed for each system, and ensuring it travels with content throughout the supply chain — are the ones that realize the full value of that investment.

Clara Aler, head of marketing, Knox Media Hub: Modern MAM systems are designed to address the complexities of metadata management; they are even built around metadata to structure catalogs and drive automation. We often serve to harmonize metadata between systems, production deliveries and distribution requirements. However, many organizations still struggle with disconnected standards and manual processes, such as passing CSV files between teams via email.

Advertisement

Álvaro Montalbán, chief sales officer, Knox Media Hub: The metadata journey is often quite erratic and many organizations struggle to have a proper organization, especially when automatic generated metadata comes into play. In order to make the most of the different metadata sources and maximize the quality of the metadata preserved and distributed, organizations need to have clear guidelines on metadata structure within their organization.

How are organizations measuring success when modernizing workflows: speed, cost savings, scalability or something else?

Bob Caniglia, director of sales operations, Americas, Blackmagic Design: Speed is often the first and most visible indicator that a workflow modernization effort is working, particularly as organizations produce more content for streaming and digital platforms. Cost savings are important, but organizations are looking beyond simple reductions in hardware or facility expenses. In many cases, the greatest savings come from improving efficiency and enabling teams to do more with the same resources. Many organizations are judging success by the quality of collaboration.

John Mailhot, SVP, product management, at Imagine Communications: It ultimately comes down to business outcomes — ROI, total cost of ownership, and how quickly new services can be deployed. Broadcasters are looking for systems that reduce friction, improve human efficiency, and make it easier to launch channels or adapt workflows without starting from scratch. If the technology isn’t clearly supporting revenue growth or cost control, it’s not really solving the problem.

Clara Aler, head of marketing, Knox Media Hub: While cost savings remain the primary driver, we also see organizations measuring throughput per team to see how much media they can effectively distribute across various channels. Additionally, many are embracing SaaS models that align costs with the temporality of their content.

Álvaro Montalbán, chief sales officer, Knox Media Hub: Mostly efficiency, as a broader term for cost saving. Also because it applies to a wider spectrum of areas: productivity from staff, previous workflows replacement, technical resources usage, etc…

How are companies approaching version control and content lifecycle management across multiple platforms and formats?

Guillaume Aubuchon, VP, product management, Avid: Version control has moved well beyond simple file naming conventions. Organizations need systems that understand editorial relationships across multiple tools and platforms: versions, iterations, derivatives, and rights. Media‑aware storage and collaboration platforms play a critical role here. Maintaining deterministic performance while tracking how assets evolve has become even more important. When combined with centralized metadata services, this allows teams to manage the full content lifecycle, even as work spans multiple formats and destinations.

Brett Beers, chief architect and innovation, TMT Insights: Version control has become significantly more complex as content is repurposed across regions, formats, and rights windows. Organizations are focusing on maintaining a clear, authoritative view of assets and their relationships, with metadata playing a critical role in tracking how content evolves over time. Without that foundation, it becomes very difficult to manage scale or avoid redundant work.

Rich Zabel, VP, media supply chain, Diversified: The organizations getting this right are treating lifecycle management as an architectural decision made before the first file is ingested, not a cleanup problem solved after the storage bill arrives — defining retention rules, version hierarchies, and platform-specific format relationships at the design stage so the orchestration layer can enforce them automatically throughout the asset’s life. Intelligent orchestration tooling is the differentiator here, because managing dozens of format derivatives, language versions, and platform-specific cuts manually across a growing library is not a staffing problem you can solve by adding people — the system has to understand the relationships between assets and act on business rules without human prompting. A disciplined 3-2-1 strategy — three copies, two different media types, one offsite — is now the baseline expectation rather than best practice, and organizations that haven’t formalized it are one storage failure or ransomware event away from finding out exactly what their informal approach was actually worth.

Looking ahead, what capabilities will define the next generation of media workflows?

Yaya Selva, CMO, Net Insight: The next generation will be defined less by individual tools and more by how intelligently the whole environment can be governed. The capabilities that matter most will be cross-domain orchestration, service-level visibility, trusted interconnection, policy-driven automation and predictable behavior across hybrid, multi-vendor and multi-organization workflows. That is where operational maturity is heading.

Thomas Carlisle, senior solutions architect, TAG Video Systems: The next generation will be defined by intelligence layered over instrumentation, enabling operators to understand the service-level impact of a problem before viewers notice. We’ll also see much tighter integrations between production monitoring and business systems. Operational data will be informing decisions at the editorial and commercial level, not just the technical one.

Guillaume Aubuchon, VP, product management, Avid: Next‑generation workflows will be defined by adaptability. They’ll be context‑aware, agentic, and increasingly predictive. We will be able to anticipate what media is needed, where, and by whom. The combination of open APIs, MCP-enabled integrations, and agentic intelligence across a unified platform is what transforms a collection of tools into a competitive advantage.

Geoff Stedman, CMO, SDVI: The next step beyond workflow automation is workflow intelligence — the ability to make operational decisions dynamically, based on real-time data, rather than executing predefined rules. This means supply chain platforms that can route work based on content value, adjust resource provisioning as conditions change, and surface the operational data that lets teams continuously improve rather than just maintain. Organizations that are starting to instrument their supply chains now — capturing cost and performance data at every step — are building the foundation for this kind of intelligence.

Advertisement

John Mailhot, SVP, product management, at Imagine Communications: We’re moving toward fully unified, software-based environments where linear and streaming workflows operate together rather than separately. IP infrastructure is the foundation for that, because it allows resources to be deployed and reassigned dynamically across content streams and locations. The key capability going forward is adaptability — being able to change workflows in real time, based on a flexible routing and management environment underneath.

Brett Beers, chief architect and innovation, TMT Insights: The next generation of workflows will be defined by how effectively organizations can connect and activate their data across the entire media supply chain. AI will play an important role, but its impact will depend on the quality and accessibility of the underlying data it relies on. We’ll also see a shift toward more exception-based operations, where automation handles the routine, and human expertise is applied where it drives the most value.

Rich Zabel, VP, media supply chain, Diversified: AI-driven logging is where the immediate transformation is happening — comprehensive ingest-time analysis of visual elements like faces, logos, objects, on-screen text, and graphics combined with full audible intelligence including transcription, translation, and both in-shot and ambient audio recognition will make today’s metadata practices look primitive, and the organizations building their MAM and orchestration strategies around that richness now will have a compounding advantage. The challenge with AI vendors specifically is that the capability curve moves faster than any other technology category in the stack — what was a differentiator six months ago is table stakes today, which means the commercial relationship and contract structure matter as much as the technology itself, because you need the flexibility to evolve without being locked into yesterday’s model. The partners worth investing in are the ones who treat the engagement as a continuous learning relationship — bringing intelligence about where the technology is heading, what’s working operationally across their customer base, and what to avoid, rather than simply selling a point-in-time capability and moving on.