header-langage
简体中文
繁體中文
English
Tiếng Việt
한국어
日本語
ภาษาไทย
Türkçe
Scan to Download the APP
Partnered With:

Founded 5 months ago, raised three rounds of funding, former Huawei chief expert Jianye Hao reeled in hundreds of millions of yuan for AI memory technology.

According to the Beating Monitor, AI memory company MemoraX AI has completed a multi-million-dollar seed ++ round of financing, led by Shangshi Capital, Camellia Capital, and Ren'ai Capital, with participation from multiple institutions and additional investment from existing shareholders. The company was founded in March of this year and completed three rounds of financing in April, May, and August.

Founder Hao Jianye previously served as the director of Huawei's Decision Reasoning Laboratory, Large Model Algorithm Laboratory, and Chief Technology Officer of the Healthcare Division, and is also an elite professor at Tianjin University. The company received tens of millions of dollars in its first round, followed by tens of millions of RMB in the second round, with the total amount from the first two rounds exceeding 100 million RMB.

MemoraX AI focuses on long-term memory for large models. It aims to train a specialized memory model using reinforcement learning, teaching AI what to remember, when to recall it, rather than relying solely on vector database storage and retrieval of historical information. Its memory system previously scored 82.65 on LoCoMo-Refined and the company has also participated in building two long-term memory evaluation sets: ScriptMem and SWE-context-bench.

The company has also launched MemoraX Code, providing cross-session long-term memory for Claude Code and Codex. It stores knowledge from code repositories, engineering experiences, and user habits, allowing the Coding Agent to pick up where it left off in a new session based on past projects.

RelatedRecommended
Defactor is still in its early days and does not break out business figures.
Cryptocurrency mining companies: "Volunteers" for blood transfusion chips, "Pursuers" for low-voltage chips, and "Saviors" for Nvidia graphics cards.
If all goes well, as early as the end of this year.
Correction/Report
Submit
Add Library
Visible to myself only
Public
Save
Choose Library
Add Library
Cancel
Finish