About the Company
Our client is a global AI technology company helping organizations address complex business challenges through deep learning, big data, and IoT solutions. Its diverse, international team collaborates across cultures to develop technology that creates meaningful impact in society.
The company works closely with major enterprises to move artificial intelligence from research into practical, reliable applications. Its culture combines technical excellence, customer focus, curiosity, and a strong commitment to continuous learning. Employees are encouraged to exchange knowledge, explore emerging technologies, and contribute through research, publications, presentations, and open-source initiatives.
This is an opportunity to join a technically ambitious organization where your work can influence both cutting-edge AI development and the way mission-critical businesses operate. You will collaborate with talented colleagues, engage directly with business stakeholders, and help transform advanced language models into solutions that deliver measurable value.
What the Job Entails
As a Data Scientist specializing in LLM development, you will drive the research, development, and practical implementation of large language models, including vision-language models. You will own the full development lifecycle, from data design and model training to evaluation, inference optimization, infrastructure development, and operational improvement.
You will research model architectures and learning methods, improve performance through pre-training, instruction tuning, alignment, and reinforcement learning, and develop capabilities such as long-context processing, tool use, and agent-based systems. You will also design benchmarks for Japanese-language and specialized business domains, monitor model quality, and address risks including hallucination, data leakage, and privacy.
Working with project managers, product leaders, and engineers, you will translate business requirements into model strategies, support commercial deployment—including on-premises environments—and establish continuous improvement cycles. You will also contribute to technical reviews, knowledge sharing, mentoring, process standardization, and a strong research and engineering culture.
Key Requirements
At least three years of experience researching and developing machine learning models.
Hands-on experience training large language models, regardless of model scale.
Strong ability to investigate model errors at the log level, develop appropriate hypotheses, and implement corrective actions.
Deep enthusiasm for continuously pursuing advances in large language model research and development.
Experience designing or operating model training, evaluation, or inference workflows.
Ability to connect technical decisions with product, customer, security, cost, and operational requirements.
Strong collaboration skills and the ability to work effectively with product managers, project managers, software engineers, and business stakeholders.
Experience with NVIDIA Megatron-LM or NeMo frameworks is advantageous.
Experience building or operating MLOps environments and working with distributed processing is advantageous.
Contributions to the field through competitions, conference presentations, technical writing, journal publications, or open-source software are valued.
Comfort participating in everyday conversations and written communication in English.