MemoraX AI, Founded by Huawei Expert in Shenzhen, Secures $10M to Solve Large Model Memory Issues

In the AI race, large language models have achieved remarkable progress in logical reasoning and multimodal interaction, yet they still suffer from an embarrassing flaw: they forget as soon as users turn away. Recently, a new startup called Shenzhen Yiji Yuan Technology Co., Ltd. (MemoraX AI) emerged, officially announcing a multi-million dollar seed round. The round was led by L2F Lightsource Venture Fund and Zhongding Capital, with participation from several well-known investors.

Notably, MemoraX AI is based in Nanshan, Shenzhen. It took less than a month from founding to securing this significant investment.

Industry Leader: A Dual-Hybrid Expert Bridging Academia and Industry

He Jianye, the leader of MemoraX AI, is a classic scholar-entrepreneur. In his 40s, he boasts an exceptionally strong academic background and conducted postdoctoral research at the Massachusetts Institute of Technology (MIT). After returning to China in 2015, he founded one of the country's earliest deep reinforcement learning laboratories at Tianjin University.

Before founding MemoraX AI, He Jianye spent many years at Huawei, where he served as director of the Decision Reasoning Laboratory, director of the Large Model Algorithm Laboratory, and Chief Technology Officer of the Medical Group. As Huawei's chief expert in decision intelligence, he led multiple billion-level industrial projects. This blend of top-tier academic insight and hands-on industry experience gave MemoraX AI a big-company DNA and a practical orientation from day one.

Technical Breakthrough: Giving AI Long-Term Memory

Currently, AI agents often struggle with unstable information storage and inaccurate retrieval, resulting in poor user experiences. MemoraX AI's core approach is to embed memory capabilities directly into the model's underlying architecture through its proprietary "Agentic RL" technology, rather than simply bolting on an external database.

According to the company, this technology addresses three key challenges:

Dynamic Evolution: Memory evolves from static data to being continuously understood, updated, and reconstructed during conversations.

Precise Recall: In relevant memory tests, its performance significantly surpasses comparable solutions, while training efficiency has improved by hundreds of times.

Scenario Reusability: Whether in daily office tasks, code writing, or digital companionship, the memory seamlessly transfers across scenarios.

Business Vision: Creating Tailored Digital Companions

The founding team believes that if AI cannot close the gap between storage and memory, it will forever remain merely an efficient search engine. MemoraX AI aims to transform AI from a cold tool into a warm companion.

The company's business landscape is already taking shape. On the B2B side, MemoraX AI plans to offer standardized memory modules for industries like finance, healthcare, and law, addressing efficiency pain points from repetitive inquiries. On the consumer side, it aims to build personalized intelligent assistants that genuinely understand user preferences, learning habits, and work needs. The first batch of standardized memory products is expected to launch within the next year.