Recently, the results of the 2024 Tianjin Science and Technology Awards were announced. KeenData successfully won the first prize in Science and Technology Progress Award for its breakthrough achievements in complex dynamic scene understanding and application in the field of intelligent visual perception. This honor not only represents high recognition of its technical capabilities, but also demonstrates its unique competitiveness centered on the Data&AI All-in-One strategy and AI-Native integrated intelligent-driven architecture, providing strong support for building a unified "Data and Intelligence Infrastructure" for the intelligent era.
It is understood that this project focuses on the complex dynamic scene understanding and application challenges in the field of intelligent visual perception, addressing key issues such as spatiotemporal inconsistency of multi-source multimodal data, image inconsistency in cross-camera fusion, and blurred spatiotemporal boundaries of complex actions. As a leading Data&AI infrastructure software provider, KeenData focuses on the national strategic direction of "Data Elements × AI", supporting the evolution of China's digital economy core systems. Based on years of technical accumulation in the Data&AI field, the company has developed more than ten industry-leading core technologies including multi-architecture unified management and multimodal data processing, effectively solving the common problems in existing visual perception systems such as insufficient functional utilization, large data volume with poor quality, serious information silos, and insufficient intelligence levels, achieving industry-leading overall technical capabilities.
Data&AI All-in-One Strategy
Currently, with the increasing complexity of enterprise business and the rapid penetration of AI technology, more and more enterprises are adopting multi-data engine and multi-instance combination strategies in data platform construction. Combined with the exponential growth of unstructured data and the widespread adoption of multi-cloud environments, the coordination difficulty of cross-environment, cross-engine, and cross-instance data management has significantly increased. The rise of AI large models has further raised higher requirements for efficient data retrieval, refined governance, and full-process collaboration.
Based on deep insights into the laws of enterprise digital transformation, KeenData proposed the "All-in-One" strategy, relying on its self-developed lakehouse data intelligence platform KeenData Lakehouse and full-stack intelligent development base Keen AI, deeply integrating data integration, storage, governance, asset management and other aspects with AI technology. Based on connecting underlying metadata, it has achieved secure sharing of multi-source heterogeneous data, efficient scheduling of model training and inference, and full-process implementation of intelligent agent development. This model not only reduces the cost of data transmission and repeated processing, but also solves the pain points of difficult multi-source data collaboration and low efficiency of AI application implementation, providing solid support for cross-domain collaboration and business innovation.
The core value of KeenData's Data&AI All-in-One strategy lies in breaking the traditional separation between data and AI technologies, using metadata management as the underlying support to achieve integrated collaboration from the full data lifecycle to the full AI process. In this model, a unified platform can complete the full-process operations from multi-source heterogeneous data aggregation and governance to model training and inference, computing power scheduling, knowledge base construction, intelligent agent development, and application implementation, truly achieving "one-stop" solutions for all aspects of the AI lifecycle: from data collection and preparation to model construction, deployment and monitoring, significantly reducing the complexity of data and AI collaboration.
In practical applications, KeenData's lakehouse data intelligence platform KeenData Lakehouse has achieved remarkable results. In the industrial sector, KeenData has built big data infrastructure and management dashboards, constructed full-link, fully visualized data development management platforms, data asset catalogs, and data synchronization systems, achieving comprehensive data management and control, reducing labor costs by 45% and improving overall efficiency by 15%. In the energy sector, KeenData has built a unified data foundation to efficiently store and compute new big data, and further combined with business scenarios to provide hundreds of service supports for planning, engineering decision-making, and engineering integration platforms, managing 15 business systems, 630+ data models, and 1200+ data tables.
Focus on AI Infra
KeenData's New AI-Native Integrated Intelligent-Driven Architecture
At the 2025 Global Digital Economy Conference, KeenData showcased its new AI Infra platform solution, the lakehouse data intelligence platform KeenData Lakehouse built on AI-Native architecture, deeply participating in data infrastructure construction in government, industry and urban sectors, empowering the high-quality development of China's digital economy.
The platform has the following features:
1. Lakehouse Storage: Provides high-performance unified storage format that can solve problems such as unstructured and structured data aggregation, data quality applications, data deletion compliance assurance, and data modification to support data capture. It endows data lakes with openness, reliability, and lifecycle management capabilities, providing efficient and secure full-lifecycle solutions for data processing and applications.
2. Multi-modal Fusion Processing Engine: As one of the core technologies of the Lakehouse platform, it has fast data processing capabilities. By optimizing query execution plans and reducing I/O operations, it significantly improves data processing performance and supports multiple data sources and query types, ensuring flexibility and efficiency in data processing.
3. Unified Metadata: Self-developed technology applied to the platform's full matrix products, enabling enterprise-level data governance, covering permission management, centralized auditing, automatic tracking of data lineage, and cross-platform, cross-tenant, cross-regional data sharing. It also ensures the security and compliance of data assets, providing advanced security solutions for data discovery and security.
4. Data Services: Provides simple and secure solutions for internal and cross-domain sharing of high-quality datasets. Enterprises can securely share high-quality result sets without copying or moving original data, reducing the complexity and cost of data transmission and repeated processing. It supports multiple data formats and protocols, ensuring broad applicability of data sharing.
5. AI Integrated Implementation Capability: Relying on KeenAI to provide high availability and low-latency inference endpoints, with the help of full-stack toolchain integration, low-code development environment, and built-in foundation large models, it efficiently implements full-domain capabilities from multimodal data processing, model training and inference, computing power scheduling to knowledge base construction, intelligent agent development, application construction, and model continuous learning optimization.
6. Full-stack Data Intelligence: In the full lifecycle of data integration, storage, development, governance, weaving, and data asset management, it deeply integrates artificial intelligence technology to achieve automation, intelligence, and efficiency in data processing, significantly improving data processing efficiency, optimizing data quality, and enhancing data value mining capabilities.
7. Multimodal Data Processing: Natively supports unified governance and management of multi-source heterogeneous data. Through unified data architecture and management tools, it improves data availability and security, better supports data and AI work, and meets the needs of cross-modal analysis scenarios such as data processing, video analysis, and document parsing.
8. Agent Intelligence: Ready to use out of the box, providing autonomous Agent factory that can quickly build Agent and RAG applications. With technological innovations such as dynamic task decomposition algorithms, multimodal intent understanding, and cross-platform execution engines, it achieves a closed loop from "thinking" to "action".
With the arrival of the digital economy era, Data & AI technology has become the core engine driving digital transformation services across all industries. KeenData, led by the "All-in-One" strategy and AI-Native approach, empowers thousands of industries to move towards a new era of digitalization and intelligence. KeenData has served leading clients including State Grid, Bank of China, Sinopec, and others, deeply participating in trusted data space construction and supporting the sustainable development of trusted data spaces. The platform's market share ranks among the top in China's data platform software and has received professional recognition from Gartner and IDC. Looking to the future, KeenData will continue to advance "Data & AI New Infrastructure" with the platform capabilities of "Trustworthy + Intelligent + Systematic", supporting government and enterprise organizations to move from "data governance" to "intelligent governance", becoming an important participant in China's new data infrastructure construction.
