"The true value of AI lies not only in technological advancement, but also in whether it can deeply penetrate the front lines of industries and solve practical problems." On December 11, Zhao Wenbiao, CEO of Ant Data, made this statement at the ecosystem partner conference, setting a pragmatic direction for the currently booming AI industry. Reporters on site learned that this AI to B service provider has delivered impressive results. Although it remained low-key for a year, it has quietly entered the leader quadrant of China's intelligent body development platforms.
Figure caption: Zhao Wenbiao, CEO of Ant Data
According to data, over the past year, Ant Data has taken "technological implementation" as its core goal, driving AI from laboratories into real business scenarios. Focusing on key sectors such as finance, energy, transportation, and manufacturing, it has partnered with nearly 20 top companies to launch more than 100 intelligent body joint solutions, truly achieving the industrial value of "customers are willing to use and continue to use."
Starting from the most challenging scenarios: Becoming the preferred choice for AI transformation in the financial industry
Customers' choices are the best evaluation. The biggest difficulty in the implementation of industrial AI often lies in the most complex and strictly regulated scenarios. Faced with industry uncertainties, Ant Data made a "contrary consensus" choice - locking the breakthrough point on the financial sector, which has the highest data barriers and strictest compliance requirements. This strategic perseverance quickly turned into market results: its services now cover 100% of state-owned shareholding banks, over 60% of local commercial banks, and hundreds of financial institutions, becoming the preferred partner for AI transformation in the financial industry.
Banks such as Ningbo Bank, Shanghai Bank, Xinhua Life Insurance, Fubon Bank, and Kunshan Rural Commercial Bank are building their own "AI brain" based on Ant Data's self-developed financial reasoning large model. At Ningbo Bank, the intelligent decision-making system co-built by both parties using Ant Data's Agentar knowledge engineering platform (KBase) has completely broken through the traditional "knowledge island" of financial services. This system realizes the transition from "fuzzy matching" to "precise reasoning" through an intelligent mechanism of "planning-retrieval-inference": the accuracy of answering complex questions has increased from 68% to 91%, the response speed has been improved to hundreds of milliseconds, and the accuracy and recall rate of recommended content have increased by 35% and 40% respectively. More importantly, the system can clearly present the reasoning process and data sources, achieving AI "black box whiteization" in the highly regulated financial scenario, making intelligent decisions both efficient and trustworthy.
The AI mobile banking application launched by Shanghai Bank is another typical case of Ant Data's technical empowerment. This application is supported by Ant Data's LUI technology, intelligent body development, and model management capabilities, with "conversational service" as its core. Users can handle high-frequency tasks such as transfers, financial consulting, and pension inquiries through natural language interaction, achieving an experience revolution from "people seeking services" to "services seeking people". Combined with personalized recommendations and age-friendly design, this AI mobile banking has become a typical sample of industry digital transformation.
Scenario cross-border extension: AI technology services for all industries and people's livelihood
The successful practice in the financial sector has accumulated rich industrial AI experience for Ant Data, and also driven its intelligent body capabilities to extend to more people's livelihood and public service areas, achieving a value shift from "financial level" to "industrial level". In Nanjing, Ant Data has jointly developed the "Xiaolanjing" public transport intelligent body with Nanjing Bus, completing the upgrade of public transport scheduling from "experience-based decision-making" to "intelligent planning".
This bus network optimization intelligent body can accurately map the city's travel pattern, analyze the passenger flow rules within 200 meters of hubs, and precisely identify the "capillary blockage points" in travel. Based on its intelligent suggestions, Nanjing Bus has opened more than 30 micro-circulation routes and added 84 new stops. Among them, the 210 route connecting the Metro Line 7 has a daily maximum passenger volume of 2,168 people, with nearly 50% of the senior citizen cards used, efficiently serving commuting populations while effectively covering nearby tourist attractions, achieving a win-win situation between social value and operational efficiency. Both parties have now launched the second phase of upgrades. The newly upgraded intelligent body can complete passenger flow prediction through natural language instructions and output visualized route planning and economic feasibility reports, making public transport scheduling more efficient and citizens' travel more predictable.
In the new energy sector, Ant Data's AI exploration has also achieved breakthroughs. The EnergyTS energy power time series large model released in March has supported the large-scale application of the "Energy Service Intelligent Body", covering the entire scenario of new energy asset investment decision-making, power trading assistance, energy system operation optimization, and equipment maintenance management. According to data, its investment decision-making efficiency has exceeded 60 times that of manual methods, and the operating and trading effects have improved by more than 10%. The asset valuation efficiency has increased by more than 10 times. It has already helped companies like Xiayin Nengke and Huopu achieve digital transformation.
Full-stack technology infrastructure ensures support, helping global enterprises upgrade to the AI era
Technology is the foundation of Ant Data's service to customers. Unlike the technical roadmap that pursues general ranking, Ant Data has always focused on three core issues: business growth, experience innovation, and risk control, building a full-stack product matrix of "large models-platform-applications". In the middle, the Finova Bench financial-specific evaluation benchmark ensures model effectiveness, and at the bottom, the Taicang Annotation Center provides data support. Combined with the cutting-edge technology exploration of the Tianji Lab, a complete closed-loop from technology research and development to scenario implementation has been formed.
This engineering capability has been verified by authoritative professionals: its data analysis intelligent body Agentar-Scale-SQL has won the first place in both accuracy and execution efficiency on the global authoritative benchmark BIRD-Bench, surpassing teams such as Google and setting the best historical performance for domestic companies in this field. Notably, BIRD-Bench covers 33GB of complex data across 37 industries. Achieving "dominance" in such benchmarks exactly proves the high adaptability of Ant Data's AI technology to the actual needs of the industry. According to the official website, since its ranking on September 25th this year, Agentar-Scale-SQL has remained at the top for over two months, leading globally. Reporters learned that the vertical model supporting this intelligent body has been open-sourced by Ant Data.
In IDC's recent "IDC MarketScape: 2025 China Intelligent Body Development Platform Vendor Assessment", Ant Data successfully entered the highest level "Leader" quadrant, thanks to its "full-stack technology capabilities" and "deep accumulation in the financial field".
At the same time, Ant Data is promoting its proven AI solutions globally: in October this year, it was selected into the Hong Kong Monetary Authority's Generative AI Sandbox Project, providing AI intelligent body services and AI security products for the sandbox, helping to improve the operational efficiency, user experience, and risk control capabilities of banks in Hong Kong. Currently, Ant Data serves over 100 overseas financial institutions, including Nan Yang Commercial Bank and Standard Chartered Bank.
"Xinglan Plan" Upgrade: Building an AI Ecosystem with 300 Partners
Ecosystem prosperity is the key to the large-scale implementation of technology. In 2025, Ant Data has established deep cooperation with 300 partners, jointly serving more than 13,000 end customers. Based on the annual practice and partner feedback, Ant Data officially upgraded its partner plan at the conference, launching the "Xinglan Plan", which enhances partner capabilities from four dimensions: technical empowerment, operational support, opportunity sharing, and financial support, promoting the overall upgrading of the ecosystem.
As a solid comrade-in-arms, the support from ecosystem partners is closely related to Ant Data's innovation in commercialization models. A partner representative said, "Ant Data has mature and leading product capabilities and brand influence. We hope to build a methodology for AI empowerment in vertical scenarios through cooperation with Ant Data, and jointly promote the large-scale implementation of industrial AI." Another ecosystem partner shared the results of the cooperation on-site. This year, 30% of its revenue comes from the business with Ant Data, and based on this, the partner has clearly expressed a strong intention to deepen the cooperation.
