In an era of surging digital waves, every technological leap begins with a profound reshaping of infrastructure.

On March 26, as a key component of the 2026 ZGC Forum Annual Conference, the "ASEAN Innovation Cooperation and Development Forum," co-hosted by ZGC Group, China Galaxy Securities, and United Overseas Bank (UOB), was held in Beijing. The forum, themed "Harnessing Innovation Ecosystems, Forging ASEAN's Future," saw Yu Yang, Chairman of KeenData Technology, invited to speak. In his keynote, "AI Data Infrastructure: The Critical Leap for Artificial Intelligence," he systematically elaborated on the pivotal role of AI data infrastructure in the large-scale implementation of artificial intelligence, starting from the core logic of AI industrialization, and outlined KeenData's strategic for the ASEAN market.

Yu Yang, Chairman of KeenData Technology, delivering a keynote speech

Opportunities lost are gone forever; those seized are never missed.

Global AI investment is currently experiencing explosive growth. Gartner predicts total global AI spending will reach $2.52 trillion in 2026, a 44% increase year-over-year. However, high investment has not yielded expected returns – NTT DATA research shows over 70% of generative AI projects fall short of expectations; a MIT report further indicates that over 95% of generative AI investments have yet to achieve financial returns.

In his speech, Yu Yang pinpointed the core issue, stating that the problem lies not with the models themselves, but with "the AI data infrastructure supporting AI operations not being ready."

Using the recently notable open-source project "OpenClaw" (related to lobster farming) as an example, he highlighted the fundamental difference between personal AI and organizational AI. Personal AI addresses individual efficiency, with data residing on personal computers, ready for immediate connection. In contrast, enterprises, governments, and large organizations face a complex. Data is scattered across dozens of business systems, requiring integration, governance, security, and compliance before it can be fed to AI.

"Preparing data for AI is not a one-time labeling task, but a systematic engineering system of AI data infrastructure," Yu Yang stated. "Above the iceberg are algorithms and computing power; beneath the iceberg are data governance, quality assurance, security, and compliance. These three elements together form the complete picture for AI implementation. Without a solid AI data infrastructure, computing power and algorithms cannot be transformed into tangible productivity."

Industry poses the questions, technology provides the answers. During the recently concluded Two Sessions (National People's Congress and Chinese Political Consultative Conference), a compelling metaphor for technological innovation resonated widely – "stretching and advancing the '1'" – used to describe how research transition from labs to the industrial frontline. For AI data infrastructure, the same principle applies: "stretching the '1'" – transforming dispersed data resources into usable AI assets, ultimately becoming fuel driving AI.

KeenData Technology, with its self-developed KeenData Lakehouse as its core, has built an AI-native Data & AI platform. This platform is driven by an AI-in-Lakehouse akehouse engine, integrating full-chain capabilities such as data integration, data preparation, data virtualization, multimodal management, model training and inference, and Agent development and orchestration.

"With this systematic engineering approach, developing application software becomes unnecessary. All intelligent implementation for an enterprise will be accomplished within this platform, intelligently completing the development of Agents across various industries through," Yu Yang explained.

Leveraging this original technological capability, KeenData Technology has received numerous: First Prize for Provincial/Ministerial Science and Technology Progress, Gartner Recommended Global Data Infrastructure Provider, IDC Recommended Lakehouse Technology Provider, Top Market Share in Advanced Technology Lakehouse Data Platforms, and National Key "Little Giant" Enterprise Specializing in Niche Sectors. Technologically, the company holds over 200 technology invention patents, with a code self-development rate of 97%.

"The value of AI data is realized in the vast array of application scenarios – on the front lines, at the grassroots level, and in small and medium-sized enterprises," Yu Yang emphasized during his speech.

KeenData Technology has already achieved large-scale validation across multiple industries, including energy, finance, retail, and urban governance.

In the energy sector, they helped clients integrate data across the entire chain of production, transmission, and consumption, enabling a leap from "resource-driven" to "AI data intelligence-driven" operations. The accuracy and timeliness of dispatch decisions significantly improved, while providing compliant and auditable data support for emerging businesses like green electricity trading and carbon footprint tracking.

In the financial sector, by connecting dispersed systems and eliminating information silos, they support the development and iteration of AI-driven intelligent risk control models. Risk identification accuracy notably increased, supporting the comprehensive implementation of AI Agents in quantitative trading, intelligent investment research, and intelligent investment advisory.

In the consumer internet sector, they omni-channel data to form comprehensive consumer portraits, enabling optimized global supply chain allocation. The intelligent supply chain management system significantly improved inventory turnover rates, while personalized recommendations based on consumer behavior data enhanced customer satisfaction and repurchase rates.

In the urban governance sector, through multi-source data aggregation and cross-departmental sharing, they a modern governance system characterized by and coordinated, enhancing thelevel of urban management and emergency response capabilities.

"Particularly noteworthy is that the value of AI data is realized through cross-industry data integration. Data, as a factor of production, and AI data infrastructure, as a production tool, enable socialized allocation, providing technology and AI data supply to more third parties. A supply model that is 'usable but invisible' is the most efficient for stimulating AI economic potential and a crucial part of large-scale AI implementation," Yu Yang said.

Continuing the journey without resting on laurels.

Standing at the beginning of the "15th Five-Year Plan," innovation cooperation between China and ASEAN is facing new historical opportunities.

ASEAN, as one of the world's most dynamic digital economies, possesses a very solid digital economic foundation in consumer internet and fintech. Its data value is full of vitality, and it stands at a critical juncture for an intelligent leap.

Currently, KeenData is undertaking data infrastructure projects in Malaysia, the Philippines, and Indonesia: a collaboration with a local group in Malaysia has entered the POC (Proof of Concept) validation stage; providing a tailored lakehouse architecture solution for a major retail chain in the Philippines; and promoting full-chain data application scenarios in the oil and gas industry in partnership with a local system integrator in Indonesia.

Seizing the moment to set sail. At the forum, Yu Yang announced KeenData Technology's three major directions for cooperation in the ASEAN market:

• First, Technical Cooperation– Establishing the first AI Data Infrastructure Overseas Service Center in Singapore to provide localized deployment, compliance consulting, and joint testing environments for Chinese and ASEAN local enterprises, significantly lowering the barrier to AI implementation.
• Second, Open Platform Capabilities– Leveraging open platform capabilities to broadly connect existing intelligent agent capabilities from China, collaborating with partners to build industry-specific AI data sandboxes. Together, they will refine reusable data governance templates and intelligent agent applications in sectors like manufacturing, energy, and healthcare, adapting them for rapid local implementation.
• Third, Collaboration with Capital– Partnering with investment institutions focused on the Southeast Asian market to launch lightweight AI data infrastructure pilot packages, supporting early-stage innovative companies in quickly validating cross-border AI scenarios and getting the technology operational first.

The tallest tree grows from a tiny sprout; the highest tower rises from a pile of earth.

In the conclusion of his speech, Yu Yang stated: "The critical leap for AI industrialization lies in the synergy of computing power, algorithms, and data. Computing power is the engine, algorithms are the soul, and data is the fuel – without a high-quality supply of AI data, even the strongest computing power and best algorithms cannot be truly implemented. The role of AI data infrastructure is to help every enterprise, every industry, and every sector transform scattered data resources into usable AI data assets, enabling the creation of their own value and achieving an intelligent leap. We look forward to joining hands with all partners here to empower the high-quality development of ASEAN's digital economy with AI data infrastructure and collectively create a more prosperous future."

From serving leading domestic enterprises to empowering ASEAN's digital economy development, from responding to national strategies to participating in global technological cooperation, KeenData Technology is using AI data infrastructure as a lever to write a new chapter for Chinese technology companies 'going global'. This gathering in spring the powerful forces of unity and progress. As the of data elements converge into the mighty rivers of the AI economy, and as Chinese wisdom deeply integrates with ASEAN vitality, a more prosperous and resilient regional AI digital ecosystem is coming into being.