On January 20, 2026, during the Alibaba Cloud PolarDB Developer Conference, Li Feifei, who serves as the Senior Vice President of Alibaba Cloud and also heads the Database Products Division, made an announcement. He stated that PolarDB has rolled out new features, including the AI Data Lakehouse (Lakebase), which deeply integrates large - scale model capabilities into databases. This integration empowers data systems to not only store and query multimodal data but also directly facilitate AI - driven intelligent decision - making.

Li Feifei highlighted that AI - native databases are an inevitable trend in the progression of technology. In recent times, memory prices have seen a sharp increase, soaring by 30% to 40%. Moreover, there is a potential for these prices to surge two to three folds in the future. Given this situation, it is of utmost importance to expedite the AI - driven transformation of databases.

He stressed that large models that are disconnected from hot data are likely to produce inaccurate or nonsensical outputs (a phenomenon known as "hallucinations" in the AI field) and are unable to meet the demands of real - time scenarios. The key to bridging the gap between hot data and model inference lies in model operatorization. This approach enables the real - time conversion of hot data into Tokens. At present, Alibaba Cloud PolarDB has deployed over 3 million cores, spanning 86 availability zones across the globe.