Welcome to our latest edition of the Investing in AI paid version that analyzes specific public market stocks from a bottoms up AI approach. We ask who can benefit or suffer from the trends in AI, and write up that perspective. Today we cover Box. Disclosure: We do not own any position in Box. We are AI people, not financial advisors, so please do your own research before making a decision.
I. Executive Summary
Box is often perceived as a mature cloud storage company competing with Dropbox or SharePoint. In reality, it has evolved into an intelligent content platform positioned at the control plane of enterprise AI. The file is no longer the product—it is the sensor.
Every document stored in Box feeds a growing corpus of unstructured enterprise data. That data powers AI models, enables agentic workflows, and supports high-margin automation use cases across regulated industries. This creates the Box flywheel: content generates data, data powers AI, and AI drives deeper platform adoption and monetization.
The legacy storage business provides revenue stability, but the long-term upside lies in Box’s Intelligent Content Management (ICM) platform and emerging agentic AI layer. From a bottom-up AI perspective, Box is positioning itself as infrastructure—not an app—for enterprise AI workflows. This transition matters now, as enterprises shift from AI experimentation toward governed, production-scale deployment.
II. The Bottom-Up AI Architecture
The Raw Material: Unstructured Enterprise Data
Roughly 80–90% of enterprise data is unstructured—PDFs, contracts, presentations, emails, scanned documents, and multimedia files. While most AI narratives focus on models, Box’s strategic advantage begins one layer lower: data custody.