Multilingual AI Researcher - Data & Models

Confidential; United Kingdom; On-site; GBP 70,000 - 120,000; Full time

Benefits offered by this job

Competitive compensation

Equity stake

Publish opportunities

Open-source collaboration

Job summary

United States Digital Space LLC, on behalf of a partner, seeks an AI Researcher – Multilingual Data in the United Kingdom to lead the development of multilingual datasets and research strategies that power state-of-the-art language models across diverse languages and domains.

You will bridge research and engineering, publish findings at leading AI conferences, contribute to open-source initiatives, and translate research outcomes into production-ready improvements in a fast-moving startup

Qualifications

• Advanced background in NLP/ML/AI or related field.
• Proven research experience in multilingual or cross-lingual language modeling with publications.

Responsibilities

• Design and conduct research on multilingual datasets (collection, filtering, deduplication, quality assessment).
• Develop strategies for low-resource languages via sampling, data augmentation, and curriculum learning.
• Improve multilingual LLMs via cross-lingual transfer, alignment, and robust representations.
• Build and refine multilingual evaluation benchmarks across languages.
• Collaborate with ML engineers to influence training pipelines and deployment.
• Publish findings at AI/NLP conferences and contribute to open-source projects.

Skills

Multilingual NLP

Data collection & curation

Research publication

Python

Curriculum learning

Low-resource languages

Tools

PyTorch

JAX

Job description

United States Digital Space LLC, on behalf of a partner, seeks an AI Researcher – Multilingual Data in the United Kingdom to lead the development of multilingual datasets and research strategies that power state-of-the-art language models across diverse languages and domains.

You will bridge research and engineering, publish findings at leading AI conferences, contribute to open-source initiatives, and translate research outcomes into production-ready improvements in a fast-moving startup