Smartbi's 15th Anniversary: From Enterprise BI to BI+AI, Launching Baize AgentBI
On the occasion of its 15th anniversary, Smartbi has extended its enterprise BI capabilities into a BI+AI platform and launched Baize AgentBI, the Baize Ontology Intelligence Platform, the Baize Knowledge and Experience Management Platform, and the Baize Multi-Agent Collaboration Platform. The company says it is exploring multi-agent collaboration on an metrics foundation, achieving a 10x improvement in R&D delivery efficiency, and ranked first in the comprehensive score across seven platform technology capabilities in IDC's "China GenBI Vendor Technology Capability Assessment, 2025." At the current stage, most enterprises are suited to start with SmartBI Insight and Baize AgentBI.
Body · AI translation
“Can this number be trusted?”
Behind a single number or a single conclusion lie a company's data systems, metric framework, and business logic. What SmartBI has done over fifteen years is make these links trustworthy.
Deeply rooted in business: fifteen years of accumulated enterprise-grade BI capabilities
01 | Beyond reports: a deep understanding of companies' real business
SmartBI was founded in 2011. From the very beginning, what we set out to do with “Smart” was intelligent and advanced analytics.
At its founding, SmartBI already served financial institutions and leading companies across industries. These large customers have extremely high requirements for data accuracy, stability, and security. Rich experience with complex projects and solid product capabilities have given us the privilege of participating in data-related work for multiple Chinese aerospace projects.
For many people, BI is often understood as stopping at reports, dashboards, and data visualization. But in SmartBI's view, these are only the presentation forms of data analysis. The real value of BI lies in going deep into the business, understanding the business relationships and management logic behind the data, and making data truly serve business operations and decision-making.
In serving leading enterprises, SmartBI has developed three mutually supporting enterprise-grade capabilities:
1 Metrics and data foundation. From data integration and metric definition to calculation logic, data relationships, and permission management, we establish a unified data and metric system and build a solid data foundation.
2 Business models and industry knowledge. We codify business objects, operating rules, analytical methods, and industry know-how into business models, so that data can correspond to real business scenarios.
3 Professional enterprise-grade delivery capability. We integrate product capabilities into complex enterprise environments and accumulate experience through implementation, service, operations, and continuous iteration, truly adapting to business changes.
02 | Beyond traditional analytics: continuously exploring new paradigms of intelligence
Long before large models emerged, SmartBI had already begun exploring natural language analytics. In 2017, we launched “Smart Xiaomai,” enabling users to query data and obtain analysis results through natural language. Since then, SmartBI has continued to deepen natural language analytics (NLA) technology and obtained an invention patent authorization for it in 2021.
As enterprises' data management needs continued to deepen, SmartBI further made the development of a metrics system an important foundation of its BI platform. In 2023, it became the first to launch a metrics management platform, driving deep integration of metrics management and BI analysis, so that unified metric definitions, calculation logic, and business models run through the entire analysis process.
In 2024, SmartBI became the first in China to deeply integrate AI Agents with BI, launching SmartBI AIChat Baize, so that AI is no longer limited to simple Q&A but gradually participates in more complete data analysis tasks.
In 2025, SmartBI continues to deepen the AgentBI concept, advancing intelligent analytics from single agents to multi-agent collaboration and exploring how more complex analysis tasks can be completed through collaboration among different agents.
Running through this entire technological evolution has always been an understanding of enterprises' business needs and a continuous effort to bring data analytics closer to the business. This has also earned industry recognition: in IDC's “China GenBI Vendor Technical Capability Assessment, 2025,” SmartBI ranked first in the comprehensive score across seven platform technical capability dimensions and received full marks in the two industry capability dimensions of finance and central state-owned enterprises.
Over the past fifteen years, SmartBI has taken a steady, down-to-earth path, insisting on long-term service and going deep into and thoroughly understanding each customer's business.
Entering the AI era, large models and agents have also become users of data. But general-purpose large models do not understand each individual enterprise; and when departments build their own agents and knowledge bases separately, data, metrics, and permissions become fragmented, forming new “AI islands.”
01 | From data analysis applications to a trusted supply of data and business cognition
The metrics systems, data models, business rules, and industry knowledge that SmartBI has continuously accumulated over the past fifteen years are precisely the important foundation for building this business base.
With BI as the foundation and AI as the engine, SmartBI continues to deepen the building of an enterprise-grade BI+AI platform and to construct an enterprise-grade trusted data platform that AI can call upon. BI is responsible for accumulating and governing enterprises' trusted data and business cognition, while AI enables these capabilities to be called by different models and agents for analysis and business tasks.
Activate existing BI assets, connect existing data assets across different business systems, promote reuse of existing assets, and reduce duplicate construction.
02 | Connecting underlying assets with upper-layer applications to build a full-chain enterprise intelligence
Having clarified the capabilities required for enterprise AI applications, how to implement them quickly and effectively around each enterprise's own characteristics is the next focus. Over the past fifteen years, in the process of continuously serving leading enterprises and solving complex business problems, SmartBI has continuously accumulated data construction and application practices and gradually developed a methodology for enterprise intelligent data construction.
This methodology is oriented toward business value and, around enterprises' actual needs, consolidates the data foundation, unifies metric definitions, gradually improves business models and knowledge systems, and integrates AI capabilities into business applications. During construction, it insists on prioritizing key scenarios and taking small, quick steps, relying on the platform to reuse construction results, integrating data operations into daily business, and driving data capabilities to continuously release value.
Around this process, SmartBI does not merely deliver software; rather, with “platform + solutions + services + operations,” it provides customers with full-chain support, accompanying enterprises as they gradually build their own data-driven and intelligent capabilities and driving the true transformation of data value into business value.
03 | Building an open, continuously evolving enterprise BI+AI capability system
Based on enterprises' varying construction needs from data foundations to intelligent applications, SmartBI has built a new-generation AI product architecture that connects enterprise data, business knowledge, intelligent analytics, and multi-agent collaboration, forming an open, continuously evolving enterprise AI capability system.
Enterprise intelligence does not start from zero. For most enterprises, data integration, modeling, metric system construction, data governance, and analytical applications remain the focus of current digitalization efforts. Through SmartBI Insight, SmartBI connects enterprise databases and business systems, integrates scattered structured data, establishes unified data models, metric definitions, and permission systems, and through reports, dashboards, and self-service analytics, transforms data into information that is understandable, analyzable, and usable for business decision-making. From data aggregation and unified management to analytical applications, SmartBI addresses the question of how enterprises can truly make good use of data, and it is also the core capability for enterprises to establish a trusted data foundation.
BI has established the governance and analysis foundation for enterprise structured data. For AI to further understand enterprise business, it is also necessary to systematically organize the business semantics, rules, and professional experience behind the data. The Baize Ontology Intelligence Platform sorts out business objects, terminology, relationships, rules, and advanced algorithms to help AI understand the business meaning behind data; the Baize Knowledge and Experience Management Platform carries out Know-how knowledge extraction, extracting professional knowledge from documents, meeting minutes, and other materials, establishing tags and associations, solving unstructured problems, and enabling enterprise experience to be retrieved and invoked.
With the support of data and business context, intelligent analytics can further move toward task execution. The Baize Agent Decision Analysis Platform (Baize AgentBI) is responsible for analysis execution, tool invocation, decision parsing, and result delivery; the Baize Multi-Agent Collaboration Platform organizes different agents to collaboratively complete complex tasks through task decomposition, agent scheduling, and process governance. Enterprises can also connect other agents and external AI tools via API, CLI, and other methods, and initiate tasks through natural language, business system workflows, APIs, or enterprise IM and other entry points.
This architecture supports enterprises in carrying out construction progressively according to their own data foundations and business needs. At the current stage, most enterprises are suited to starting with SmartBI Insight and Baize AgentBI, solidifying the data governance and analysis foundation, and truly using AI to unlock value. As the construction stage develops, they can then gradually introduce capabilities such as ontology intelligence, knowledge and experience management, and multi-agent collaboration, progressively improving the enterprise AI capability system.
In practice, SmartBI uses the metric foundation as a basis to explore multi-agent collaboration: humans clarify goals, formulate execution paths, and control key nodes, while agents undertake specific tasks and divide labor and collaborate, driving changes in R&D and delivery methods. This practice has achieved a 10x improvement in R&D delivery efficiency and received IDC's “Mastering the AI Supercycle: Best Practices in China's AI-Driven Business Innovation” honor.
Fifteen years, rooting deep and growing far
Over the past fifteen years, SmartBI started from enterprise BI and has consistently adhered to technological innovation and serving large customers. It is precisely this development DNA that has given SmartBI broader development space and more comprehensive acceleration in the AI era. Moving from BI to BI+AI is not following the trend, but a natural connection between existing technological accumulation and the development needs of the AI era, allowing SmartBI to naturally move faster and more steadily in this era.
Build trusted data, unlock business value.
This article is from the WeChat public account “SmartBI,” author: chosen by 6,000+ large customers, published by 36Kr with authorization.