On January 20th, the Alibaba Cloud PolarDB Developer Conference commenced at the Shanghai International Convention Center. Set against a tech-themed backdrop of code and data, Alibaba Cloud unveiled a suite of new AI capabilities, including an AI Data Lakehouse, Model Operatorization, and managed services for Agent application development. This technological leap is revolutionizing data processing for over 20,000 enterprise users globally. From financial risk management to autonomous driving, from government operations to AI-driven entertainment, Alibaba Cloud PolarDB is reshaping the data infrastructure of the future with its seamlessly integrated AI functionalities.

Li Feifei, Senior Vice President of Alibaba Cloud, charted a clear course for the evolution of database technology in his keynote address: “From cloud-native to AI-ready, and ultimately to AI-native, PolarDB stands as a robust ‘intelligent data engine,’ continuously deepening the synergy and innovation between AI and databases.”

This integration is not a mere overlay of databases and AI but a profound fusion and reconfiguration. Databases are transitioning from passive data repositories to active, intelligent decision-making engines.

In traditional AI application development, data must be extracted from the database, processed through complex ETL (Extract, Transform, Load) workflows, and then fed into specialized AI systems—a laborious and convoluted process. In contrast, ‘AI-native’ databases directly embed AI capabilities at the data storage layer, enabling systems to not only store and query multimodal data but also to directly fuel intelligent decision-making.

Alibaba Cloud outlines four foundational pillars for this new breed of database:

• Multimodal AI Data Lakehouse
• Efficient Integrated Search Capabilities
• Model Operatorization Services
• Backend Services for Agent Application Development

These pillars collectively form the bedrock of an ‘AI-ready’ database, offering users a smarter, more efficient, and secure AI data infrastructure.

Currently, Alibaba Cloud PolarDB’s deployment spans over 3 million cores, covering 86 availability zones worldwide. This expansive infrastructure network provides unwavering support for its AI-native transformation.

The AI Data Lakehouse stands as a cornerstone of PolarDB’s innovation. Designed for unified lake-warehouse integration, it seamlessly merges the flexibility of data lakes with the high performance of data warehouses through unified storage and efficient analytics. This solution dismantles traditional data silos, ensuring consistency and unified access for full-modal data.

The AI Data Lakehouse offers scenario-specific caching acceleration, providing IO and bandwidth boosts for various scenarios to ensure efficient data flow. Its multimodal engine, deeply integrated with unique In-DB model operatorization, enables developers to perform semantic retrieval and inferential processing directly within the database.

Model operatorization technology marks another significant breakthrough in Alibaba Cloud PolarDB. Through this technology, AI models can be directly incorporated into database query workflows, akin to traditional SQL operations.

Within the database, it supports in-database inference, Agent-Ready architecture, and AI long/short-term memory mechanisms, empowering databases to not only store and query data but also to directly drive intelligent inference and decision-making.

This architectural innovation ensures that data processing and AI inference can be completed within the same environment, eliminating security risks and processing delays associated with data migration between disparate systems.

The AI-native PolarDB is no longer a theoretical notion; it is being actively implemented in the core systems of multiple industries worldwide.

Luo Bin, Senior Data Architect at Li Auto, shared their practical experience: “PolarDB’s numerous AI-designed capabilities, such as KVCache and Supabase, enable us to swiftly develop AI scenario-based applications.”

In Li Auto’s scenarios involving massive real-time data processing, AI coding, and efficient retrieval of enterprise knowledge bases, PolarDB delivers robust performance support and flexible scalability.

Lu Changqing, Head of AI Interactive Entertainment Technology at Duxiaoman, provided a case study from the niche sector of AI emotional companionship. He stated, “Leveraging PolarDB AI’s KVCache inference acceleration, long-memory vector indexing, and memory management solutions, we have effectively enhanced the memory retrieval and recall efficiency of emotional companionship Agents, boosting emotional coherence.”

Notably, these capabilities significantly lower the barrier to AI application development. PolarDB accelerates value exploration for Agents in vertical industries by providing integrated services for Agent application development through Supabase multi-tenancy and Serverless packaging.

This series of enterprise practices underscores that AI-native databases are transitioning from technological concepts to practical productivity. They are lowering the technical barrier for enterprise AI application development while enhancing system performance and security.

References:

http://field.10jqka.com.cn/20260120/c674151972.shtml

https://finance.eastmoney.com/a/202601203624484795.html