In this lightning-round session, an Assistant Vice President for Research Technology demonstrates how to leverage Box AI to streamline research workflows. Unlike open-web AI tools, this platform allows researchers to create secure, custom "Hubs" to analyze specific collections of documents.
The presentation walks through the setup process—from creating a Hub to indexing files—and showcases how to use AI for targeted literature reviews and data analysis within a private, controlled environment.
Key Highlights:
Building AI Hubs: A step-by-step guide on creating custom repositories by uploading files and allowing them to index, creating a closed ecosystem for the AI to query.
Model Flexibility: Users can choose between various top-tier AI models (such as OpenAI, Gemini, or Claude) to run prompts against their data, allowing for comparative analysis of results.
Targeted Literature Reviews: A demonstration of how to command the AI to act as an academic researcher, summarizing key themes from a specific set of uploaded articles without interference from external internet sources.
Citations & Transparency: The tool provides "based on" references, linking generated answers directly to the specific source documents within the Hub for easy verification.
Data Privacy & Copyright: Important reminders about adhering to usage rights and institutional policies when uploading third-party or sensitive materials for AI analysis.