Design a Box and ChatGPT Integration for Governed Enterprise Content
Company: Box
Role: Product Manager
Category: Product Design & Strategy
Difficulty: medium
Interview Round: Other
You are a product manager at Box. Design an integration between Box and ChatGPT.
```hint Decide where the user is working
Pick the moment of need first: someone working in ChatGPT who needs company files, or someone in Box who wants a model's help with them. That choice decides which product owns the experience and what Box has to protect along the way.
```
### Clarifying Questions
- Is the goal to bring Box files into ChatGPT, to bring ChatGPT's capabilities into Box's own interface, or both?
- What does Box want most from this: keeping and expanding enterprise customers, reaching new users, or new revenue from AI features?
- Which customers are in scope: individuals on personal plans, or companies whose IT admins manage both Box and a business ChatGPT plan?
- Are some files never allowed to leave Box, such as regulated or confidentially classified content?
### What a Strong Answer Covers
- Why the integration matters to Box strategically, including the risk that users work in ChatGPT and treat Box as background storage
- A chosen segment and the concrete jobs the integration serves, such as finding, summarizing and drafting from company files
- An end-to-end flow that enforces Box permissions per user and gives admins control over data, retention and audit
- A prioritized first release versus later scope, with adoption metrics and security guardrails
- Trade-offs such as live retrieval versus copying an index, and how the integration sits alongside Box's own AI features
### Follow-up Questions
- A user loses access to a folder while a ChatGPT conversation still holds content from it. What should happen?
- Usage is high, but many users stop opening Box itself. Is that a win or a loss for Box, and what would you measure?
- How would you package or price the integration?
Overview: A Box product design prompt about integrating the cloud content platform with ChatGPT. It tests strategic reasoning about where enterprise users work, choosing a segment and use cases, designing a flow that respects Box permissions and governance, and defining adoption metrics and security guardrails.