- Yokohama, Yokohama, Japan
Lead the post‑training of large language models to create domain‑specific agents for customer support and network deployment in telecommunications.
Responsibilities
- Design and execute supervised fine‑tuning pipelines for domain knowledge injection
- Apply parameter‑efficient fine‑tuning techniques such as LoRA, QLoRA, and adapter‑based methods
- Collaborate with AI & Domain Specialists to ensure high‑quality datasets
- Maintain a structured catalog of fine‑tuned model variants and training configurations
- Design and implement policy‑reward training pipelines using GRPO and SAPO