Alibaba’s Qwen Office aims to automate spreadsheets, chats, reports, and PPTs

卡尔的AI沃茨 · wechat · 2026-07-28

This long WeChat post tests Alibaba’s new Qwen Office across several real office workflows and argues that AI will enter daily life through repetitive work automation, not just flashy demos.

What it was tested on

E-commerce analysis: The author used Qwen Office to pull product, supplier, pricing, and review data from multiple platforms, then turn it into a CSV and analyze negative reviews, sentiment, trends, and business risks. The result reportedly identified issues much faster than manual review.
Market research: It was asked to compare Qwen, Doubao, Tencent, and Microsoft Copilot in the Chinese AI office market, then produce a report, a comparison table, and an HTML PPT with charts, visuals, sources, and no hallucinated data.
Chat ops / weekly summaries: The tool was fed exported WeChat group chats in Markdown and used to summarize daily discussion points, decisions, responsibilities, and generate an HTML weekly report.
Ops analytics: It was also used to turn scattered social metrics into a weekly report that could be pushed to DingTalk on a schedule.

Main takeaway

The post’s core argument is that the real “AI-native” moment is when agents quietly take over boring work: cleaning Excel, reading chats, compiling reports, and generating deliverables you can actually ship.

Original post →

Alibaba’s Qwen Office aims to automate spreadsheets, chats, reports, and PPTs

More from coding & agent