About the Role
Claude is used by millions of people across the globe. Making it work exceptionally well for all of them, regardless of what language they speak, including low-resource and underserved languages, is one of the highest-impact opportunities at Anthropic. It’s how we expand Claude’s reach to non-English-speaking communities, nonprofits, schools, and organizations that stand to benefit the most. We’re looking for a Research Technical Program Manager to own this effort. You’ll be responsible for improving the quality and language coverage of Claude, and ensuring its responses are locally and culturally relevant. This means working across research, product engineering, beneficial deployments, and external partners to define what world-class multilingual performance looks like, build the programs and evaluation frameworks to get us there, and ship improvements that users worldwide can experience. This role is ideal for someone with deep NLP and linguistics expertise who is energized by the challenge of making AI genuinely useful for communities beyond well-resourced, English-dominant markets.
Responsibilities:
- Own the end-to-end program for improving Claude’s multilingual capabilities, from identifying language-specific gaps through research execution to production
- Partner with pre-training and post-training research teams on multilingual data strategies and training approaches to improve Claude’s performance across languages
- Design and implement evaluation frameworks for multilingual model quality, including benchmarks for low-resource languages and regression testing across model updates
- Coordinate multilingual dataset collection and curation, working with internal teams and external partners to build high-quality training and evaluation data across languages
- Act as connective tissue between research, product engineering, and data infrastructure teams to reduce chaos and accelerate multilingual improvements
- Manage strategic partnerships with external organizations for dataset acquisition, linguistic expertise, and multilingual evaluation, including technical scoping, data quality requirements, and integration workflows
You May Be a Good Fit If You:
- Have a background in NLP, computational linguistics, or multilingual AI research
- Have several years of experience in technical program management, with a track record of driving complex programs that span research and engineering teams
- Can engage credibly with researchers on multilingual training data strategies, evaluation methodology, and model capability assessment. You’re not just tracking tasks, you’re contributing to technical direction
- Have experience building evaluation pipelines and frameworks that give meaningful signal about model quality across languages, not just benchmark scores
- Have worked with multilingual datasets at scale and understand annotation frameworks across languages, data quality challenges, and the operational complexity of sourcing data for languages with limited digital resources
- Have experience coordinating with external data partners, linguistic consultants, or research institutions for dataset creation or evaluation across languages
- Understand how research improvements translate into shipped product features and can work across research and product engineering to move multilingual capabilities from prototype to production
- Build trust quickly with research teams. Researchers see you as someone who amplifies their impact, not someone who adds process overhead
- Bonus if you speak multiple languages and have seen firsthand how language technology performs across them
Deadline to Apply: None, applications will be received on a rolling basis.
