DEEPEXI TECH collaborates with Huawei Cloud to launch a data intelligence solution, adding "cloud power" to accelerate the scaling of enterprise intelligent bodies

On August 12, DEEPEXI TECH and Huawei Cloud jointly released a data intelligence solution based on DEEPEXI TECH's DeepWorks enterprise intelligent agent platform and FastData enterprise-level data intelligence solution. This solution relies on Huawei Cloud's technological foundation and targets key industries such as intelligent manufacturing and retail, covering critical scenarios including production operations, supply chain collaboration, equipment maintenance, quality management, and store operations. By connecting enterprise data, business systems, and intelligent agent capabilities, it forms a closed loop from data insights, business decisions to execution feedback, helping enterprises enhance operational insights, decision-making responsiveness, and resource allocation efficiency, promoting lean operational management, efficient organizational collaboration, and continuous improvement in human efficiency, while accelerating product, service, and business model innovation with Data+AI.

In the joint solution, DeepWorks is responsible for agent orchestration and task execution, serving as one of the core capabilities of the solution. It integrates capabilities based on AgentArts, including Harness, multi-layer memory, tool invocation, and trusted evaluation, and undertakes the knowledge foundation capabilities of DEEPEXI's enterprise large model and over 2,000 industry Skills accumulated by Deepology, transforming enterprise knowledge, business experience, and industry capabilities into callable, executable, and sustainably iterative capabilities for intelligent decision-making and automated execution in complex business scenarios.

FastData provides the data engineering and data governance foundation for intelligent agent applications. By launching a command-line interface (CLI) and packaging standardized Skills into an Agent Kit, FastData extends data platform capabilities from the web console to scenarios such as Agent, IDE, and CLI. Enterprises can use natural language, with DeepWorks as the entry point, to drive FastData to complete tasks such as data source management, metadata governance, data modeling, SQL execution, and task operations, promoting the transition of data platforms from "manual operation" to "agent execution."

For DEEPEXI TECH, the release of this joint solution marks that DeepWorks, based on Huawei Cloud's technological foundation, provides enterprise customers with a more flexible, secure, and stable operating environment. Leveraging Huawei Cloud's advantages in computing power, cloud-native, security compliance, and industry service systems, DeepWorks and FastData can operate collaboratively under a unified cloud service framework, further enhancing the delivery efficiency and replicability of the solution across multiple industries and scenarios.

Currently, intelligent agents are transitioning from technological exploration to standardized application and large-scale implementation. With the issuance of relevant documents such as "Implementation Opinions on the Standardized Application and Innovative Development of Intelligent Agents" and "Artificial Intelligence Intelligent Agent Interconnection Part 1: Overall Architecture," enterprises' demand for intelligent agent platforms has evolved from merely general-purpose building tools to platforms capable of deeply integrating into business scenarios, connecting enterprise data and systems, accumulating organizational experience, and supporting long-term operations.

The differentiated value of DEEPEXI TECH's DeepWorks is reflected in this. It is not a general-purpose agent aimed at a single office scenario but an intelligent agent platform built for real business scenarios of enterprises Relying on the continuous accumulation of Deepexi's self-developed enterprise large model in key industry ontology paradigms, data governance, and enterprise knowledge bases, DeepWorks is able to access and understand the specific knowledge semantic environment of enterprises, transforming enterprise data, business processes, tool systems, and industry experience into callable, executable, and sustainably iterative intelligent capabilities. On this basis, DeepWorks can complete more accurate demand understanding, task planning, and execution collaboration around the actual needs of enterprises, supporting intelligent decision-making, process automation, and application construction in complex business scenarios, helping enterprises upgrade AI from an office assistance tool to an execution engine focused on core business.

Currently, Deepexi has served over 400 enterprises and has formed a capability chain covering data governance, knowledge modeling, and intelligent application in fields such as manufacturing, retail, scientific research, and public utilities. In scenarios such as manufacturing and leading retail, the task execution accuracy of DeepWorks has consistently exceeded 95%. Meanwhile, the success rate of industrial intelligent tasks in actual industrial engineering deployments has reached 97.8%. In terms of performance, in the first half of 2026, the company achieved operating revenue of 284 million yuan, a year-on-year increase of 115%; among which, AI business revenue increased by 209.2% year-on-year to 226 million yuan, with gross profit increasing by 120.5% year-on-year to 160 million yuan, and the quality of earnings continues to improve, with revenue scale, business structure, and profitability optimizing simultaneously.

Market analysts say that the data intelligence solution jointly launched by Deepexi and Huawei Cloud not only further strengthens DeepWorks' capabilities in cloud delivery, business collaboration, and scenario replication but also provides a more practically valuable path for enterprise intelligent agents to move from "usable" to "easy to use and frequently used." As the joint solution is implemented in more industry scenarios, Deepexi's differentiated advantages in the enterprise intelligent platform track are expected to be further amplified