Cua Verified
Research Intern (Summer 2026)
San Francisco, California, United StatesRemoteInternshipJunior / Entry-level$8,000–$9,252 /yrPosted 2 months agoHidden Gem · YC Startup
Role summary
Cua is seeking a Research Intern for Summer 2026 to work on the infrastructure for general-purpose AI agents. The intern will prototype, test, and benchmark multi-modal LLM-based agents, collaborating with engineers and researchers. Responsibilities include generating multi-modal data, designing agent systems, automating benchmarking, exploring training techniques, and developing evaluation tools. Qualifications include a PhD student status in Computer Science, experience in applied research, familiarity with modern agents, and hands-on experience with PyTorch, Python, and cloud compute. Preferred qualifications include experience with reinforcement learning, open-source contributions, and benchmark design. This is a paid internship with a flexible, remote-friendly setup.
## Overview
Cua is building the infrastructure that enables general-purpose AI agents to safely and scalably use real computers and applications.
We're a small team backed by **Y Combinator** and top-tier investors, and our open-source tools are already used by thousands of developers. As a **Research Intern**, you’ll help prototype, test, and benchmark multi-modal LLM-based agents - from data pipelines to orchestration systems.
You’ll collaborate with engineers and researchers to turn cutting-edge ideas into real systems and benchmarks that can be shared with the community. This is a chance to contribute to open-source research, design experiments, and explore the frontiers of agentic AI.
---
## Responsibilities
* Generate and curate large-scale, high-quality multi-modal data (GUIs, browsers, system UIs)
* Design and test single- and multi-agent systems for data and computer use
* Automate benchmarking of agent orchestration (with or without human-in-the-loop)
* Explore new training and inference techniques to boost reasoning and action-taking (e.g., RL-based agents)
* Develop benchmarks, tools, and datasets to evaluate agentic capabilities on Cua
* Collaborate with the founding team and contribute to research publications, open-source tools, and the broader community
---
## Qualifications
### Required:
* Currently a **PhD student** in Computer Science or related field (strong Master’s considered)
* Experience in applied research with a solid publication record
* Familiarity with modern multi-modal or reasoning agents (e.g., OS-Atlas, Qwen, GUI-R1)
* Hands-on experience with **PyTorch**, **Python**, and cloud compute (AWS, GCP, etc.)
* Comfortable designing experiments, evaluating models, and working with multi-modal data
* Excited by generative AI, agent systems, and pushing the boundaries of what’s possible
### Preferred:
* Experience with reinforcement learning or agent-based training methods
* Prior contributions to open-source projects or benchmark design
* Familiarity with large-scale dataset construction and evaluation pipelines
* Interest in bridging research and engineering for real-world applications
* Based in or able to spend time in **SF/Bay Area** (preferred), but remote OK
---
## What We Offer
* **Research impact** – Opportunity to publish, open-source, and influence open agent research
* **Hands-on projects** – Work directly with engineers and researchers on cutting-edge systems
* **Open-source visibility** – Contribute benchmarks and datasets used by the community
* **Flexible setup** – Remote-friendly; SF-based team
* **Learning environment** – Collaborate on projects at the intersection of infrastructure and AI research
---
## How to Apply
Please include:
* Your **CV** and GitHub/portfolio
* A short note on a research problem you’d like to tackle
* Bonus: try building something with [Cua](https://github.com/trycua/cua) or suggest a benchmark idea — we notice contributors
---
**This is a paid internship (3-month full-time preferred; part-time considered). Compensation will depend on location and experience.**
**Cua AI, Inc. is committed to fair and transparent opportunities. We encourage applicants from all backgrounds, identities, and walks of life to apply.**
**Personal data will be handled in accordance with the GDPR (EU Regulation 2016/679) and other applicable data privacy laws.**
Cua is building the infrastructure that enables general-purpose AI agents to safely and scalably use real computers and applications.
We're a small team backed by **Y Combinator** and top-tier investors, and our open-source tools are already used by thousands of developers. As a **Research Intern**, you’ll help prototype, test, and benchmark multi-modal LLM-based agents - from data pipelines to orchestration systems.
You’ll collaborate with engineers and researchers to turn cutting-edge ideas into real systems and benchmarks that can be shared with the community. This is a chance to contribute to open-source research, design experiments, and explore the frontiers of agentic AI.
---
## Responsibilities
* Generate and curate large-scale, high-quality multi-modal data (GUIs, browsers, system UIs)
* Design and test single- and multi-agent systems for data and computer use
* Automate benchmarking of agent orchestration (with or without human-in-the-loop)
* Explore new training and inference techniques to boost reasoning and action-taking (e.g., RL-based agents)
* Develop benchmarks, tools, and datasets to evaluate agentic capabilities on Cua
* Collaborate with the founding team and contribute to research publications, open-source tools, and the broader community
---
## Qualifications
### Required:
* Currently a **PhD student** in Computer Science or related field (strong Master’s considered)
* Experience in applied research with a solid publication record
* Familiarity with modern multi-modal or reasoning agents (e.g., OS-Atlas, Qwen, GUI-R1)
* Hands-on experience with **PyTorch**, **Python**, and cloud compute (AWS, GCP, etc.)
* Comfortable designing experiments, evaluating models, and working with multi-modal data
* Excited by generative AI, agent systems, and pushing the boundaries of what’s possible
### Preferred:
* Experience with reinforcement learning or agent-based training methods
* Prior contributions to open-source projects or benchmark design
* Familiarity with large-scale dataset construction and evaluation pipelines
* Interest in bridging research and engineering for real-world applications
* Based in or able to spend time in **SF/Bay Area** (preferred), but remote OK
---
## What We Offer
* **Research impact** – Opportunity to publish, open-source, and influence open agent research
* **Hands-on projects** – Work directly with engineers and researchers on cutting-edge systems
* **Open-source visibility** – Contribute benchmarks and datasets used by the community
* **Flexible setup** – Remote-friendly; SF-based team
* **Learning environment** – Collaborate on projects at the intersection of infrastructure and AI research
---
## How to Apply
Please include:
* Your **CV** and GitHub/portfolio
* A short note on a research problem you’d like to tackle
* Bonus: try building something with [Cua](https://github.com/trycua/cua) or suggest a benchmark idea — we notice contributors
---
**This is a paid internship (3-month full-time preferred; part-time considered). Compensation will depend on location and experience.**
**Cua AI, Inc. is committed to fair and transparent opportunities. We encourage applicants from all backgrounds, identities, and walks of life to apply.**
**Personal data will be handled in accordance with the GDPR (EU Regulation 2016/679) and other applicable data privacy laws.**
