Autonomous AI: Agentic production looks to further revolutionise broadcasting

By Shirish Nadkarni

Sophisticated agentic production with the help of artificial intelligence (AI) is being increasingly employed in the broadcasting arena. It uses autonomous AI systems to plan, coordinate and execute complex media workflows from concept to distribution. 

Unlike traditional generative AI that creates single pieces of content on demand, agentic systems use multi-agent teams to reason through multi-step goals, share context, and make micro-decisions.

“AI in broadcasting automates workflows, enhances audio-video quality, and scales content creation across live and on-demand media,” said Sandeep Dutta, Amazon Web Services (AWS) president for India & South Asia.

“Basically, an agentic product is an autonomous software application powered by AI that can independently plan, make decisions, and execute multi-step tasks to reach a specific goal with minimal human supervision.

“An agentic product has agency; it can use external tools, call APIs, remember past context, and adapt its approach as conditions change. These systems have the ability to act independently, but in a goal-driven manner.”

Dutta cited the example of a travel planner, which is a tool where you type “Book a weekend trip to Kulu-Manali under INR 40,000 (about US$420)”, and it checks flights, selects a hotel, coordinates dates, and buys the tickets, using your permission.”

“AI agents can communicate with each other and other software systems to automate existing business processes,” added Dutta. “But beyond static automation, they make independent contextual decisions. They learn from their environment and adapt to changing conditions, enabling them to perform sophisticated workflows with accuracy.”

Agentic AI can be single or multi-agent setups.

In a single-agentic AI system, one AI agent handles all tasks sequentially. These are preferable when businesses need a faster solution that can work on a well-defined problem or process.

Multi-agentic AI, on the other hand, involves multiple AI agents collaborating to break down complex workflows into smaller segments. This approach is more scalable than single systems and is much more flexible for solving complex scenarios. The vast majority of agentic AI agents refer to this latter, more diverse form of AI deployment.

“Amazon Bedrock is a fully managed service that offers a choice of industry-leading foundation models (FMs), along with a broad set of capabilities needed to build generative AI applications,” said Dutta. 

Amazon Bedrock Agents use the reasoning of FMs, APIs, and data to break down user requests, gather relevant information, and efficiently perform tasks. Building an agent is straight-forward and fast, with setup in just a few steps. It supports memory retention for task continuity, and multi-agent collaboration to build multiple specialised agents under the co-ordination of a supervisor agent.

“We have introduced an open-source toolkit with a growing catalogue of starter agents purpose-built for healthcare and life sciences use cases,” Dutta says. “AWS Transform is the first agentic AI service for transforming .NET, mainframe, and VMware workloads. 

“Built on 19 years of migration experience, it deploys specialised AI agents to automate complex tasks like assessments, code analysis, refactoring, decomposition, dependency mapping, validation, and transformation planning.

“It helps organisations to simultaneously modernise hundreds of applications while maintaining quality and control.”

Amazon Q Business is a generative AI-powered assistant designed to help a user find information, gain insights, and take action at work. It puts the power of agentic AI creation in the hands of every employee. Anyone can use it to create lightweight agentic AI apps that interact with common enterprise software and automate repetitive tasks.

Another company that is demonstrating practical AI applications that add value both today and long-term is Imagine Communications.

Imagine is showcasing AI capabilities across its ad tech and video infrastructure portfolios, demonstrating practical AI innovation that can help media organisations reduce costs, surface new revenue opportunities, and automate critical workflows without compromising the security and reliability that their operations depend on.

Steve Reynolds, CEO, Imagine Communications, said, “We are approaching AI from an operational and workflow-centric perspective through a structured three-element framework:

• Embedding operational assistive machine learning (ML) capabilities where we create immediate, measurable value;
• Introducing Model Context Protocol (MCP) interfaces that give AI systems secure, governed access to Imagine platforms; and
• Longer-term development of AI agents that can interact with media workflows in auditable ways, ensuring the trusted core products customers have relied on for decades remain a stable foundation.

“What we’re showing is not a concept. It is working technology built on the systems our customers already know and trust. Every AI demo we have is a practical application that enables media organisations to move faster, work smarter, and generate real profitability without introducing new sources of risk.”

Within Imagine’s ad tech portfolio, AI is being applied across the advertising transaction lifecycle to help sales teams work more efficiently, sharpen revenue forecasting and yield management, streamline traffic operations, and enable more dynamic commercial decision-making — while preserving the integrity of the systems of record that underpin broadcast revenue.

“We have Landmark Sales AI Assistant, which is an embedded AI assistant that lets sales teams create, copy, and amend campaigns in flight using natural language, freeing sellers to focus on high-value sales activity,” said Reynolds.

“We also have Landmark Sales MCP/ AI-Powered Total TV Trading, where buyers transact campaigns through a simple portal and conversation, while sellers manage revenue through AI-powered summaries. 

“Then there are Landmark Sales AI Dashboards, which are AI-driven in-app reporting delivered in a token-less environment with no per-query AI costs; and Landmark Rights & Scheduling – AI-assisted Channel Scheduling, with embedded machine learning that applies existing business rules at speed and scale, shortening scheduling cycles and freeing teams for higher-value editorial decisions.”

Across its video infrastructure portfolio, Imagine is applying AI to improve visibility within complex media environments, accelerate responses to operational issues, and enable operators to manage routine tasks through natural interaction rather than manual navigation of intricate workflows.

“Our Magellan Control System MCP is an AI-enabled routing and processing control built on the proven Magellan core,” Reynolds added. “By enabling plain-language user interaction to stage and execute complex operations, this MCP allows agentic systems to implement assistive actions across the production environment.”

The company’s AI-Augmented Playout Management involves conversational multichannel playout operations, allowing users to schedule bulk updates, real-time operational queries, and intelligent gap-fill from the content pool — all through a governed MCP service layer that enforces guardrails between the AI and playout automation.

Reynolds adds, “In live production, trust is everything. AI must be able to explain what it is doing, operate within defined boundaries, and work alongside operators. That’s the philosophy behind our Model Context Protocol (MCP) approach.”

“We believe the MCP interface in front of the video control plane will revolutionise the customer control of their networks.  At Imagine, we believe we are significantly ahead of the competition in terms of AI understanding and execution.”

Yet another major player in the crowded agentic production space is Avid, which calls itself the power behind the content that touches virtually every person on the planet.

After launching Avid Content Core earlier this year, Avid is now transforming production to help media organisations unite content, agentic workflows and business insights.

Wellford Dillard, CEO, Avid, said, “Across news and post production, Avid’s latest innovations will help global teams accelerate their work while preserving creative control and editorial trust.

“We are showcasing new media intelligence capabilities within Avid Content Core, connecting media across storage environments and giving agents the context to act inside trusted workflows. Avid is applying agentic media experiences across its news production, editing and audio portfolio to help teams create, collaborate, and deliver content more efficiently.”

Spanning news and post production, Avid’s latest innovations will help global teams work faster and more efficiently while preserving creative control and editorial trust.

As media organisations continue to modernise operations while managing tighter budgets and increasing content demands, Avid is embedding enterprise-grade media intelligence and governed automation across its trusted ecosystem, enabling customers to speed up routine processes and gain greater visibility into the value of their content. 

At the heart of Avid’s offerings is Avid Content Core, which, a year after its debut, continues to evolve as the intelligence layer connecting the company’s agentic ecosystem across news and post production.

Avid is introducing native asset management capabilities that connect media across customers’ own cloud and on-premises environments, including Avid NEXIS, AWS, Google Cloud, Microsoft Azure and other third-party storage, facilitating the ability to index, search, manage, and work with content wherever it resides. 

Beyond a traditional media asset management (MAM), this connected media understanding provides the foundation for agentic automation to execute work within trusted production workflows, without requiring everything to be migrated into a single repository.

“It is exciting to see Avid Content Core expand and address the real opportunity of connecting the creative and operational sides of media with a clearer picture of how each piece of content is produced – and the value it creates,” says Dillard. “Avid Content Core is building that bridge, giving organisations the intelligence to understand how content is being created and the resources being used.”

For radio and television, specialised AI agents can take on distinct production roles. Script agents generate content tailored to specific programme formats, audience demographics and time slots, while adaptation agents transform content across formats, such as turning a long-form documentary script into radio segments or social media clips. Audience agents analyse viewer and listener data to recommend content adjustments, while quality agents evaluate output against editorial standards and brand guidelines.

In broadcast agentic production, the director’s role assumes major importance. Producers direct AI agent teams rather than doing every task manually, enabling faster iteration and more creative experimentation than traditional workflows allow.

Whether producing a learning module or a broadcast segment, the same core principles apply – define clear objectives, train specialised agents through rubrics (a title or set of instructions in a book), and maintain quality through systematic review. This makes Agentic Production uniquely versatile across the entire media landscape.

QED … carpe diem!