Achievements and Learnings of Dropbox's 2025 Summer Interns

Dropbox welcomed 43 interns to its Camp Dropbox Intern Program in summer 2025. They came from 27 universities, including 6 international institutions in Canada, Poland, and Ireland. Among them, 28 worked on the Engineering team, spending 12 weeks on hands-on projects, learning, and experiencing the Virtual First culture together.

The internship program was designed around growth and collaboration, and participants received 6,000+ hours of 1:1 mentoring. Projects were assigned to align with team and company goals, and many tasks connected to Dropbox Dash's AI features. Virtual First events, ERG activities, and the in-person Emerging Talent Summit were added on top of this, strengthening community and networking as well.

Each intern left meaningful improvements in their respective areas.

• Filesystem Data: Refined Dropbox's file history tracking system to improve testability and design quality, while reducing metadata infrastructure costs.
• ML Platform: Built AI Sentinel to monitor ML model deployment status, improving real-time visibility and deployment confidence.
• Storage Core: Added a storage status cache and degraded volume exclusion feature to reduce Magic Pocket's front end latency.
• Web Developer Experience: Built an AI-based code migration tool on top of the internal migration platform, supporting automatic PR generation and CLI execution.
• Connector Platform: Enabled faster use of the latest data from the Dash persistence store, improving downstream model training and external metadata integration.
• Retrieval Platform: Expanded USP to 20+ languages, introducing multilingual search and a language detection pipeline.
• Metrics: Developed anomaly detection reflecting seasonality in Vortex2, reducing alert fatigue.
• Find & Discover: Implemented a document preview experience within Dash, allowing file contents to be checked without context switching.
• Analytics Platform: Optimized large-scale Databricks queries and ETL, and prepared a liquid clustering transition for 500TB of mobile event logs.
• Conversational AI: Built a scalable AI web-automation agent connected to Dropbox backend APIs such as searchFile and uploadFile.

Overall, this internship was not simply an experience but led to actual improvements in Dropbox's core systems—Dash, search, storage, ML operations, and data platforms. Building on this experience, the company continues to recruit talent to build innovative products and infrastructure together.