Glossary

LLM Fine-Tuning

الضبط الدقيق لنماذج اللغة الكبيرة

Training a large language model further on a company's own dataset to improve performance on specific tasks—such as generating proposals in a particular tone, classifying leads by Gulf region, or answering customer questions using a brand's actual service list. Fine-tuning costs more than RAG but produces a model that inherently 'knows' the business without needing documents at runtime. Best suited for high-volume, repetitive AI tasks in Gulf enterprise deployments.