Recently, the 8th South Information Conference was grandly held in Guangzhou under the theme “The AI Era: Opportunities and Challenges for CIOs”. The event brought together industry leaders and technology innovators to explore pathways and breakthroughs for enterprise digital transformation amid the rising wave of AI. As an enterprise-level leader in Data and AI technologies, KeenData was invited to attend and showcase its technological achievements and extensive industry practices in the field of AI infrastructure (AI Infra). At the conference, Yin Zhiguo, General Manager of KeenData South China, delivered a keynote speech offering an in-depth interpretation of how a Data and AI integrated system is becoming the core engine of enterprise digital transformation in the AI era. KeenData was also honored as a 2025 CIO Preferred Digital Service Provider.

In the digital economy, data has become the fifth key factor of production—alongside land, labor, capital and technology—with its amplifying, multiplying and compounding effects on traditional production factors becoming increasingly prominent.

At the national level, China continues to increase investment in data infrastructure development: In 2021, the 14th Five-Year Plan explicitly emphasized “unlocking the potential of data elements”. In 2022, the 14th Five-Year Plan for Digital Economy Development set the goal of preliminarily establishing a data-element market system by 2025. In 2023, the National Data Bureau was established to coordinate the integration, development and utilization of national data resources. Driven by both policy momentum and technological innovation, enterprise digital transformation is entering a new phase of “deep AI penetration”.

In his keynote speech titled “Strengthening the Data Foundation as the Prerequisite for Building Enterprise AI Applications — Constructing a New Generation of Data and AI Integrated Intelligent Infrastructure Based on the KeenData Product Matrix”, Yin Zhiguo pointed out that the rapid advancement of AI technologies is reshaping the logic of enterprise data processing—from “data supporting business operations” to “AI activating data value”. However, the large-scale implementation of AI applications depends on high-quality and scenario-specific “data fuel”, and the data foundation serves as the core engine that provides this fuel.

“If AI applications are likened to fast-moving cars, then the data foundation is the stable chassis and the efficient fuel-supply system,” he emphasized. The greatest challenge enterprises face today, he noted, is handling the more than 80 percent of data that is unstructured—including audio, video and images—and enabling seamless end-to-end flow of data from collection and processing to AI model training and inference.

Looking back at the evolution of enterprise data-processing technologies, data infrastructure has undergone three major upgrades:

• Traditional IT era: Centered on relational databases such as Oracle, focusing on structured data and relying on “single-machine + siloed architecture”, which could only support gigabyte-level batch processing.
• Cloud computing and big data era: Semi-structured and unstructured data exceeded 50 percent. The rise of the Hadoop ecosystem and cloud-native architectures enabled petabyte-level distributed processing, yet data storage and AI applications remained fragmented.
• AI era: Unstructured data surpasses 80 percent, forcing architectural evolution toward the Lakehouse model, which requires unified multimodal data storage, high-dimensional matrix computation, integrated batch–stream processing and, critically, AI-native capabilities that seamlessly connect data governance with model training and inference.

KeenData’s KeenData Lakehouse platform is a Data and AI integrated intelligent foundation designed specifically for the needs of the AI era. The system is built around Three Unifications:

• Unified Format: Enables standardized storage for multimodal data.
• Unified Processin: Supports coordinated real-time and batch computing.
• Unified Catalog: Ensures full-lifecycle management of data assets.

By integrating capabilities across the entire chain—from data integration, preprocessing, analytics and exploration to machine learning, generative AI and data governance—KeenData Lakehouse establishes an All-in-One closed-loop system, transforming data from fragmented silos into intelligent, high-value assets.

Against the backdrop of the marketization of data elements, enterprise digital transformation is no longer only a technical matter but requires the creation of a highly collaborative and efficient ecosystem. According to Yin Zhiguo, KeenData aligns its approach with national policies and industry development trends, adopting the overall framework of “build, serve, govern and operate” and using its foundational platform as the carrier to establish a data-space management and operation system. This system enables participation from data providers, data users and data development service partners. Under the premise of safeguarding data rights for all parties, it promotes multi-source data integration and utilization, improves the efficiency of both data supply and data usage, facilitates bilateral and multilateral coordination among participants, supports trusted data circulation and compliant data sharing and drives the release of data-element value.

• Build: Define objectives and technical pathways and construct the foundational platform based on KeenData Lakehouse to integrate large-scale, multi-source data.
• Serve: Provide one-stop data services, establish business models and collaboration mechanisms and improve the efficiency of data supply and data usage.
• Govern: Establish unified standards and governance frameworks, ensure data security and compliance and clarify responsibilities, rights and data entitlements.
• Operate: Enable participation from multiple stakeholders, support supply and demand matching and trusted data circulation and drive the release of data-element value.

As an enterprise-level leader in Data and AI technologies, KeenData has gained broad industry recognition. The platform breaks through the traditional separation between data systems and AI systems through its pioneering AI-in-Lakehouse architecture, which unifies the lakehouse engine, OLAP-based data governance and AI technologies to create a streamlined and efficient All-in-One technical solution. The self-developed multimodal computation engine completes the entire process from data cleaning to analytical output in a single pipeline with severalfold improvement in GPU inference throughput. Combined with KMI inference acceleration, model quantization and Unity Catalog, the platform enables intelligent governance across multiple modalities

KeenData is recognized as a national-level Specialized and Innovative “Little Giant” enterprise and has received the First Prize of Provincial and Ministerial Science and Technology Progress Awards. The company has also earned strong endorsements from industry authorities. It was selected by Gartner as a recommended vendor for data foundation platforms and by IDC as a recommended vendor for lakehouse technologies. Its products have obtained the Trusted and Excellent Software Certification from the Fifth Research Institute of the Ministry of Industry and Information Technology, and the company demonstrates deep technical strength with more than 170 Data and AI invention patents.

Addressing the central theme of “Opportunities and Challenges for CIOs in the AI Era”, Yin Zhiguo summarized that the core mission of the CIO is to transform fragmented data resources into systematic AI capabilities. The KeenData Lakehouse platform is designed specifically to solve this challenge. By providing a unified technical foundation, it enables more efficient data governance, more practical AI application deployment and more sustainable business innovation.

To date, KeenData has supported Data and AI infrastructure construction for nearly two hundred large domestic and international organizations across more than twenty industries. These include Fortune 500 companies such as Sinopec, China Unicom, China Telecom, China CITIC Bank, AEON Group, China FAW and Sany Group, all of which have adopted the KeenData Lakehouse platform to build their Data and AI foundations.

Looking ahead, as the data element market continues to mature, KeenData will continue to advance research and development in Data and AI infrastructure software. Guided by the vision of building data infrastructure for thousands of industries and supporting the high quality development of the digital economy, the company aims to help more enterprises seize the initiative in the AI era and achieve a transition from data driven operations to intelligence led development.