From Hype to Impact: What NAB 2025 Taught Us About AI in Video

In this kickoff episode of the revived Visionular Podcast, David Lea is joined by Zoe Liu and Krishna Rao to reflect on their experience at NAB 2025. From AI innovations to meaningful product conversations, the trio unpacks what made this year different—and why Visionular’s AI-driven compression and enhancement tools are gaining real traction.

This isn’t just another recap. It’s a behind-the-scenes look at what’s really happening in the media and streaming space—from live sports to low-latency video to GPU-based encoding—and why "AI that works" is the future of our industry.

Key Talking Points   

• What stood out most at NAB 2025—quality over quantity in booth conversations
• AI fatigue vs. AI validation: real-world adoption across production, delivery, and playback
• Why GPU encoding is real now
• New challenges (and surprising demand) in live sports streaming
• The rising need for low-latency, edge-compatible, device-agnostic solutions
• Zoe Liu on 20 years of video quality work finally productizing into value
• The importance of small, lean, efficient AI models for video delivery
• NAB as a validation platform for roadmap, feedback, and strategic direction

About Visionular   

Visionular is an AI-driven video compression and streaming optimization company helping video platforms deliver higher-quality experiences at lower bitrates. With deep codec innovation and perceptual tuning, our tools are used across live sports, UGC platforms, RTC infrastructure, and PGC services.

• Learn more: https://visionular.ai/
• Follow us: Visionular on LinkedIn