Box s Journey to Agentic AI

The YouTube video features Ben Kus, CTO of Box, discussing his company's journey into AI, specifically focusing on agentic platforms. Box, an unstructured content platform for large enterprises, initially faced challenges extracting structured data from the vast amount of unstructured data their clients held. While early generative AI models improved data extraction, they struggled with complex documents, diverse file formats, language barriers, and maintaining high accuracy for enterprise needs. This led Box to adopt an agentic AI architecture, which involves orchestrating multiple AI models and tools in a directed graph to break down complex tasks, cross-reference information, and refine results, ultimately achieving the required accuracy for their enterprise customers. Kus emphasizes that this agentic approach provides a clean, evolvable architectural layer that addresses the limitations of simpler AI integrations and is crucial for developing advanced capabilities like deep content research.