We're looking for a Language Model Expert to help us develop and fine-tune models for a new AI-powered product in a complex technical domain. You’ll be working on real-world applications where model reliability, domain alignment, and iterative improvement are key. ## What You’ll Be Doing - Fine-tuning and customizing open-source or proprietary LLMs for task-specific goals - Building and optimizing RAG pipelines, prompt strategies, and custom instruction sets - Working with domain experts to align model output with real-world expectations - Evaluating model performance (accuracy, latency, hallucination control, etc.) - Supporting integration of models into a production-ready system ## Requirements ### Must-Haves - Strong understanding of LLM architecture, training, and fine-tuning workflows - Hands-on experience with tools like Hugging Face, LangChain, LlamaIndex, etc. - Proven experience with prompt design, RAG pipelines, or instruction tuning - Ability to work with complex, structured or domain-specific data - Strong experimental mindset and ability to balance research with production needs ### Nice-to-Haves - Familiarity with vector databases, retrieval strategies, or hybrid search - Experience in building LLM-powered applications or agents - Exposure to security, infrastructure, or developer tools domains - Japanese language capability (bonus, not required) ## Work Environment - Fully remote - Flexible hours (Core time: 10:00–14:00) - Laptop and monitor provided - Access to LLM experimentation tools (GPU environments, GPT APIs, etc.)