ByteDance’s workplace platform Feishu has rolled out a new AI-powered tool designed to help employees find answers from within their company’s internal knowledge base.

Called Feishu Knowledge Q&A, the feature is now in open beta via a web portal and integrates directly into the Feishu app.

The tool functions like an AI assistant trained on company-specific data. Once users define the scope of accessible information, the system can retrieve content from authorized chats, documents, wikis, and files to generate precise responses—without requiring manual uploads. All answers are permission-aware, meaning they vary based on the requester’s access level to protect sensitive information.

The product supports hybrid search, tapping both company knowledge and external sources, and leverages models like DeepSeek and ByteDance’s own Doubao.

By building on Feishu’s existing permission framework, the tool aims to solve a growing concern for companies using AI: preventing confidential data from being exposed when fed into public models. Feishu’s approach keeps queries and results within the enterprise’s own environment.

The launch follows a broader industry push to embed generative AI in workplace tools. Rivals such as Tencent’s Ima and Alibaba’s Quark have stepped up their investments in AI-powered knowledge retrieval, while startups like YouMind and Remio are also gaining traction with personal knowledge management products.

🔗 Source: 36Kr

🧠 Food for thought

1️⃣ Enterprise AI adoption moving from experimentation to specific use cases

Feishu’s Knowledge Q&A represents a shift in enterprise AI adoption patterns toward solving specific workflow challenges rather than implementing general AI technologies.

This approach addresses key adoption barriers identified in industry research, as 81% of organizations in 2019 were still evaluating AI rather than deploying it, with 23% citing difficulties in identifying clear use cases as a major obstacle 1.

The “no manual uploads or data prep” feature directly tackles another significant barrier, lack of data infrastructure, which has prevented 23% of organizations from successfully implementing AI 1.

By focusing on a specific knowledge retrieval use case that leverages existing company data, Feishu’s approach aligns with expert recommendations that successful AI adoption requires identifying the right use cases and developing a strong data foundation 2.

This targeted approach contrasts with earlier phases of enterprise AI where companies struggled to translate investment into practical applications, as evidenced by the gap between the 92% of companies planning to increase AI investments and the mere 1% considering their AI deployment mature 3.

2️⃣ Evolution from productivity tools to context-aware knowledge systems

Feishu’s new tool signals an important transition in workplace AI from general productivity enhancement to deeply contextual knowledge systems that understand organizational structure.

Recent Feishu developments