- 経験
- どれでも
- 給料
- —
- 求人情報
- 1
- 投稿済み
- 2時間前
- Work mode
- 在任中
- 教育
- Undergraduate or graduate program
- Eligibility
- Currently enrolled undergraduate or graduate students with prior exposure to retrieval or related fields are encouraged to apply. The role is intended for candidates who are interested in applied research and can contribute from a collaborative, feedback-friendly mindset.
- Resume
- Required to apply
Where you'll work
仕事内容
About the role
Mixedbread is focused on solving one of the biggest constraints in AI: search. The company is building a new generation of multimodal retrieval systems that can find the most relevant context regardless of whether the source is text, images, or another format.
As a Research Intern, you will work in an applied research setting alongside senior researchers on projects that aim to move retrieval forward. The internship is meant to be hands-on and practical, with exposure to both research development and product impact. Each project is expected to result in a concrete deliverable such as a blog post, a paper, or an open-source release.
Key responsibilities
- Take ownership of a retrieval research project and carry it forward with support from senior researchers.
- Build, test, and assess new methods for information retrieval.
- Work on multimodal search problems involving text, images, and other content types.
- Study how large language models can improve or change retrieval workflows.
- Help share research outcomes through papers, blog posts, or open-source model releases.
- Offer feedback to strengthen and refine the internship experience.
Research areas
- Late interaction approaches: developing fine-grained retrieval methods that go beyond ColBERT-style models.
- Omni-modality: creating retrievers that work naturally across different media types.
- LLM-augmented search: understanding how LLMs can generate, rank, and use retrieval outputs.
- Evaluation: building tests and benchmarks that reflect real-world agentic scenarios.
- Data work: improving retrieval signals and datasets to boost model quality.
- Interpretability: analyzing why retrieval systems perform well or poorly.
Requirements
The company is looking for candidates who are currently enrolled in an undergraduate or graduate program and already have some exposure to retrieval, embeddings, or related areas through projects, papers, blog posts, or open-source work. You should be comfortable coding in Python and able to implement models from the ground up. A genuine interest in information retrieval and its importance to AI is important, along with excitement for late interaction and newer retrieval approaches. A collaborative attitude and openness to feedback are also expected.
Compensation and benefits
This internship comes with competitive pay, though the exact stipend amount is not specified. The package also includes close mentorship from experienced researchers, health, dental, and vision coverage, visa sponsorship with relocation help, a professional development budget, access to leading AI tools and subscriptions, team off-sites and conference attendance, transportation support, wellness benefits such as gym membership and sports club subscriptions, and food support.
Equal opportunity statement
Mixedbread is an equal opportunity employer and values a diverse, inclusive workplace. Hiring decisions are made without regard to race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, or disability status.
Application details
Applicants will be asked to share standard resume and profile information, including name, email, resume, LinkedIn profile, GitHub profile, X profile, a project or achievement they are proud of, and what interests them about joining Mixedbread.