Responsibilities
- Apply advanced supervised and reinforcement learning methods such as SFT, DPO, and GRPO to fine-tune state-of-the-art language models
- Utilize extensive query and response datasets to improve model accuracy and efficiency across search, research, and discovery platforms
- Monitor emerging advancements in large language model research, particularly in training, optimization, and personalization
- Develop and deploy preference-based optimization systems to tailor user interactions
- Design proprietary enhancements that push the boundaries of existing model capabilities
- Transform theoretical research concepts into practical algorithms and deploy experimental models
- Manage end-to-end pipelines for data processing, model training, and performance evaluation
- Create scalable and maintainable training frameworks built on Megatron and PyTorch for post-training workflows
- Develop infrastructure components necessary for high-performance, large-scale model training
- Ensure smooth integration of trained models into production environments and core applications
- Collaborate with software engineers to embed models within the broader product suite
- Coordinate with cross-functional teams to deliver unified AI-driven experiences
- Partner with product teams to identify user requirements and convert insights into model refinements
Compensation
Competitive salary and equity package commensurate with experience
Work Arrangement
Hybrid or remote options available based on location
Team
Work within a fast-paced AI research and development team focused on pushing the limits of language models
Responsibilities
- Apply advanced supervised and reinforcement learning methods such as SFT, DPO, and GRPO to fine-tune state-of-the-art language models
- Utilize extensive query and response datasets to improve model accuracy and efficiency across search, research, and discovery platforms
- Monitor emerging advancements in large language model research, particularly in training, optimization, and personalization
- Develop and deploy preference-based optimization systems to tailor user interactions
- Design proprietary enhancements that push the boundaries of existing model capabilities
- Transform theoretical research concepts into practical algorithms and deploy experimental models
- Manage end-to-end pipelines for data processing, model training, and performance evaluation
- Create scalable and maintainable training frameworks built on Megatron and PyTorch for post-training workflows
- Develop infrastructure components necessary for high-performance, large-scale model training
- Ensure smooth integration of trained models into production environments and core applications
- Collaborate with software engineers to embed models within the broader product suite
- Coordinate with cross-functional teams to deliver unified AI-driven experiences
- Partner with product teams to identify user requirements and convert insights into model refinements
Visa sponsorship available for qualified candidates