PhD Systems & AI/ML Research Intern (Summer 2026)
DropboxJob Description
Role Description
Dropbox powers collaboration and cloud infrastructure for over 700 million registered users worldwide. We’re building the next generation of scalable storage, distributed compute, large-scale content processing, and secure infrastructure, with AI/ML workloads top-of-mind.
As a PhD Systems & AI/ML Research Intern, you’ll work on meaningful infrastructure problems at the intersection of distributed systems, backend services, and AI/ML workloads. Supported by senior engineers, you’ll own a project aligned with your research, that results in a working prototype with the goal of a submission for adoption internally and external publication. Come help us define and build the infra layer for our next-gen products.
Responsibilities
- Execute a defined engineering/research project aligned with your research.
- Implement a prototype that addresses a concrete infrastructure or AI-systems challenge.
- Build and run benchmarks using realistic workloads to measure performance, reliability, or security improvements.
- Document the design and implementation, including assumptions, architecture, tradeoffs, and evaluation results.
- Deliver a final artifact that includes a working prototype, reproducible results, technical write-up, and internal presentation.
- Collaborate with mentors to refine the approach, validate results, and, if appropriate, prepare findings for internal adoption or external publication.
Requirements
- PhD candidate in Computer Science or a related field graduating between Winter 2026 and Spring
More Jobs at Dropbox
View all →Senior Fullstack Product Software Engineer, Core Media & Intelligence
Dropbox
GTM Full-Stack Engineer (Python/JS) — Impact & Scale
Dropbox
CX Software Engineer
Dropbox
People Team Data Analyst, Talent Acquisition
Dropbox
Staff Data Engineer, Analytics Data Engineering
Dropbox
Want AI-powered job matching?
Upload your resume and get every job scored, your resume tailored, and hiring manager emails found - automatically.
Get Started Free