Google Translate's mission is to reduce the language barrier by translating text, speech when needed, making content more accessible to people. Our team’s main goal is to produce high quality translation models for Google Translate. These models are used by billions of users and integrated into products across Search, Cloud, Chrome, Android, Ads, and YouTube. This posting is specific for the Translate Research sub-team which focuses on improving translation and non-English capabilities within LLMs including training multilingual LLMs from scratch.
Requirements
- Experience in machine translations or related Natural Language Processing tasks.
- Experience with Machine Learning or Evaluations.
- Experience with Machine Translation or Multilingual Natural Language Processing.
- 2 years of experience in coding.
- 1 year of experience owning and initiating research agendas.
Responsibilities
- Research novel techniques for improving translation quality via modeling, data selection, and model tuning.
- Design and conduct experiments for model quality improvements, analyze results, and determine steps for moving forward.
- Develop and invent new machine translation techniques that improve the translation quality of Google Translate.
- Invent reliable automatic and human evaluation methods for machine translation outputs, and integrate findings into the machine translation model.
- setup large-scale tests and deploy promising ideas quickly and broadly, managing deadlines and deliverables while applying the latest theories to develop new and improved products, processes, or technologies.
- creating experiments and prototyping implementations to designing new architectures
Other
- PhD degree in Computer Science, a related field, or equivalent practical experience.
- One of more scientific publication submission(s) for conferences, journals, or public repositories (such as CVPR, ICCV, NeurIPS, ICML, ICLR, etc.).
- The US base salary range for this full-time position is $141,000-$202,000 + bonus + equity + benefits.
