The pace of innovation in the field of artificial intelligence is constantly accelerating. Recently, Shenzhen Yiji Yuan Technology Co., Ltd., which focuses on the "large model memory system" (hereinafter referred to as "MemoraX AI"), has once again announced a funding success. The company has completed a seed+ round financing worth tens of millions of RMB. Just about a month after its $10 million seed round financing in April, the total amount raised from two rounds of financing has now exceeded 100 million RMB. This round of financing was exclusively advised by Lightsource Capital, indicating a high level of recognition from the capital market for its potential in this vertical sector.

MemoraX AI was founded in March 2026 by Hao Jianye, former director of Huawei's Noah's Ark Lab, and has gathered top talents in the fields of reinforcement learning and artificial intelligence. The company focuses on "Agentic RL (agent-based reinforcement learning)" technology, aiming to overcome the common "memory loss" bottleneck faced by large models. Unlike traditional methods that rely on external retrieval (RAG) to obtain information, MemoraX AI tries to truly internalize memory capabilities into the model through technological innovation, achieving a deeper "intrinsic memory."

The company stated that the funds raised will be mainly invested in the deep iteration of core technologies and the large-scale engineering implementation. As the "intrinsic memory module" moves from the experimental stage to productization, MemoraX AI is accelerating the layout of core application scenarios: on the B-side, it is committed to empowering intelligent upgrades in enterprise knowledge management and professional domain knowledge bases; on the C-side, it aims to provide users with long-term and accurate memory interaction experiences by creating highly personalized smart assistants.