Kingsoft’s AI-Powered Writing Assistant
WPS Office is a productivity tool developed by Kingsoft with over 300M monthly active users worldwide. The company’s artificial intelligence (AI) department built a smart writing assistant from scratch using semantic matching algorithms such as intent recognition and text clustering. The tool helps users quickly create outlines, individual paragraphs, and entire documents by simply inputting a title and selecting up to five keywords. Kingsoft’s writing assistant uses Milvus, an open-source vector database built by Zilliz, to power its core vector processing module.
Objective
Build an intelligent writing assistant capable of generating outlines, paragraphs, and full documents.
Challenges
- Model user intentions from just a title or a few keywords. - Run millisecond-level queries on massive, unstructured text datasets. - Return consistently accurate similarity search results. - Process data in near real time.
Why verctor database
- Support for multiple mainstream indexes helped Kingsoft find the optimal configuration for an AI-powered writing assistant. - Partitioning allowed searches to be confined to specific data segments, accelerating similarity search on a corpus of ~2 million texts.
Results
- Queries completed in **0.2 seconds** on average. - Highly accurate content recommendations (**80%** average recall rate). - **40%** of content recommended to users was adopted.
"We conducted extensive market research, and Milvus stood out as a powerful and user-friendly vector similarity search engine. The platform’s flexibility makes it easy to optimize machine learning projects to better fit different scenarios."
Read the full storyshare this