Testing Smart TV and OTT applications has become one of the most complex challenges in modern video delivery. Device fragmentation, brand variations, frequent releases, and regional constraints make traditional QA approaches increasingly fragile.

During a joint webinar with Witbe, A+E Global Media shared how they transformed their QA operations using Agentic AI, real-device testing, and continuous automation; moving beyond scripted tests while keeping full human control.

This use case is based directly on the workflows, demonstrations, and production results presented during the webinar.

Watch the full recording of the webinar ↓

The challenge: scaling QA across devices, brands, and platforms

A+E Global Media operates multiple direct-to-consumer streaming services across:

• Smart TVs
• OTT platforms
• Web and mobile environments

Each brand must deliver a consistent user experience, regardless of:

• TV manufacturer or OS version
• Device generation (including older Smart TVs)
• Regional constraints (accounts, CDN, localization)

According to Anne Calabro, A+E Media’s QA director, one early bottleneck was device access. QA teams were distributed globally, with devices scattered across desks and homes, making reproducibility slow and inefficient.

But, centralizing devices was just a first step. Scaling QA across releases was the real challenge.

Centralizing real devices to test what users actually see

With Witbe, A+E Global Media consolidated their Smart TVs and devices into a shared, remotely accessible environment.

This enabled:

• Testing on real physical devices, not emulators
• Validation of UI behavior exactly as viewers experience it
• Faster reproduction of device-or OS-specific issues
• Consistent testing across brands and regions

But device centralization alone does not solve the problem of maintenance-heavy automation.

Why A+E adopted Agentic AI for QA automation

As release velocity increased, A+E Global Media needed automation that could adapt, not just execute scripts.

As explained by Kevin Keeler, Vice President of DevSecOps, QA, and Architecture at A+E Global Media, releases often land across all platforms simultaneously, so the key question became: How do you validate features continuously, without rewriting tests every time the UI changes?

Agentic AI introduced a different QA model:

• Tests are written as user requirements
• Agents interpret on-screen behavior in real time
• Navigation paths adapt automatically across devices
• Humans remain in the loop to validate and refine behavior

Instead of defining every key press, teams define what must be validated and let agents handle execution.

From scripted automation to reusable, brand-agnostic tests

One of the most visible gains shown in the webinar demos was test reuse across brands.

The same test logic:

• Runs across multiple A+E applications
• Adapts to different branding, layouts, and colors
• Executes on different Smart TV operating systems

For example, a single test can verify that:

• The Continue Watching row is present
• The correct item is highlighted
• Playback icons and progress bars match brand guidelines

This removes the need to maintain multiple scripts per brand or device.

Measurable results from production QA

The webinar deck highlighted concrete, production-level results:

• 16 releases validated across 4 platforms
• 421 features checked across more than 20 services
• 1,400+ hours of automated testing executed in just two months

These tests ran on real devices and covered:

• UI navigation
• Login and account validation
• Live and VOD content discovery
• Visual integrity and brand consistency

The result: faster validation cycles, reduced regression effort, and lower automation maintenance costs.

Continuous testing instead of release-based QA

Another key shift presented during the webinar was the move to continuous testing.

With agent-driven automation running 24/7:

• Issues are detected as soon as behavior changes
• Regressions surface after OS, backend, or UI updates
• QA teams receive feedback earlier in the delivery cycle

This complements, not replaces, API and integration testing. While backend tests validate services in isolation, Agentic AI on real devices validates the full end-to-end user experience.

A hybrid AI model designed for scale and control

Witbe also emphasized that not everything should be handled by AI.

The solution relies on a hybrid model:

• Deterministic actions (repeatable navigation) remain algorithmic
• Video quality analysis runs directly on real devices
• Agentic AI is used for decisions, assertions, recovery, and UI interpretation

This balance ensures:

• Predictable costs
• Transparent execution
• Scalable automation across large device fleets

It also preserves auditability, critical for enterprise QA teams.

What’s next: QA as a shared, cross-functional workflow

Looking ahead, A+E sees Agentic AI as a foundation for deeper collaboration:

• Earlier “shift-left” testing
• Closer alignment between product and QA teams
• Test scenarios derived directly from product requirements
• Broader adoption beyond QA into User Acceptance Testing and product validation

The long-term vision is a QA workflow that evolves as fast as the applications it validates.

Why this matters for Smart TV and OTT providers

As device fragmentation grows and release cycles accelerate, QA teams face a choice: maintain brittle scripts that tend to break often, or embrace adaptive, real-device automation built for change.

A+E Global Media’s experience shows that Agentic AI, used pragmatically and transparently, can scale QA without replacing human expertise, while delivering measurable gains in speed, coverage, and reliability.

Want to see the workflow in action?

Watch the full webinar recording below or request a demo to explore how Agentic AI and real-device automation can transform Smart TV and OTT QA at scale.