Video Intelligence Is Going Agentic

🎥 Recorded live at the MLOps World | GenAI Summit 2025 — Austin, TX (October 9, 2025) Session Title: Video Intelligence Is Going Agentic Speaker: James Le, Head of Developer Experience, TwelveLabs Talk Track: Multimodal Systems in Production Abstract: With over 90% of the world’s data in video format, traditional AI systems have struggled to extract deep understanding from dynamic visual content. In this session, James Le, Head of Developer Experience at TwelveLabs, explores how agentic video intelligence is transforming the field — enabling AI agents to reason about video, plan complex workflows, and execute sophisticated visual tasks autonomously. Drawing from real-world implementations, including MLSE’s 98% efficiency boost (reducing 16-hour highlight creation workflows to just 9 minutes), James breaks down how multimodal foundation models and agent architectures combine to unlock new capabilities in media, entertainment, and enterprise video processing. You’ll learn how to design planner-worker-reflector agent systems, manage temporal context across long video workflows, and build transparent reasoning pipelines that bridge language and visual media — pushing video AI beyond analysis into intelligent action. What you’ll learn: • Why traditional AI approaches fail to capture the complexity of video data • How to architect planner-worker-reflector agentic video systems • Implementation strategies for transparent, multimodal agent reasoning • Real-world applications driving massive productivity and ROI gains • How to identify and implement high-impact use cases in your organization.