Meituan reveals challenges and progress in enterprise AI transformation

The CEO of Meituan shared public insights about implementing artificial intelligence within the company. According to him, despite abundant availability of AI models, few organizations measure actual returns. He identified four critical misalignments: misconceptions about AI capabilities, limited application to routine tasks without affecting core operations, AI confined to business margins, and success metrics focused on immediate results rather than organizational integration. Meituan went through four phases in its transformation journey. Initially, between February and March, the company launched a broad experimentation program that resulted in excessive costs, consuming millions daily, and generating errors that harmed real operations. From April onward, it structured dedicated AI departments. Between June and July, it recognized that transformation required more than technology, integrating business, organization, and technical capability. By July, initial results began appearing in internal processes. The CatPaw platform, launched in July, represents the consolidation of these efforts. It already reaches 90 thousand employees with 30 thousand agents deployed. The system offers mobile and desktop applications with real-time synchronization, plus continuous cloud execution independent of local devices. The platform integrates Meituan's specialized knowledge in segments such as store management, customer evaluations, marketing strategies, and logistics, providing ready-to-use tools for operational diagnostics, content generation, and data analysis.