This service leverages a locally fine-tuned embedding model tailored for the Korean language to process and store user-created memo lists in a vector database. By analyzing the similarity of the vectorized data, the system ranks frequently written tasks. The ranking is further enhanced by utilizing gemma2, a local large language model (LLM), ensuring a seamless and intelligent task prioritization experience.
