NLP Research Intern
About Mimir
Mimir builds knowledge extraction infrastructure for scientific literature. Our platform lets researchers ask technical questions and get cited, verifiable answers drawn from the actual papers, figures, and data in the scientific record. We launched our public beta in March 2026 and are growing fast. Our team is small, technical, and research-heavy. Our work sits at the intersection of NLP, information retrieval, and scientific knowledge representation.
The Role
We're looking for a student to spend the summer building ML/NLP tools that help us extract structured knowledge from scientific papers at scale. You'll be designing classifiers, working with embeddings, writing prompts, and building Python tools that plug into an agentic AI backend. The problems are real and partially unsolved. How far you push is up to you.
Undergraduate, MS, and PhD students are all welcome to apply.
What we're looking for:
- Strong Python skills
- Some exposure to NLP/ML
- Familiarity with LLMs and prompt engineering
- Comfort working independently on a small team
- Curiosity about applying AI to real scientific problems
OCR, computer vision, or document parsing experience is a plus but not required.
Details:
- ~10–12 weeks, flexible on start/end dates
- $10,000–12,000 for the summer
- Remote, with a possible one-week on-site early in the summer
- We are unable to sponsor visas, but welcome applicants with existing US work authorization or international students working from abroad
Why this is a good summer:
- Work directly with two PhD founders on unsolved problems. If you're considering grad school, we can write a recommendation.
- Hands-on experience with agentic AI systems and LLM tooling in production
- Publication is not guaranteed, but is a real possibility depending on how far you push the work
- Potential to continue part-time into the fall