On December 28th, the two-day QCON Global Software Development Conference and the 15th Anniversary Conference, a top-level technology event, was successfully opened in Shanghai. With the theme of "Set Sail for AIGC Software Engineering Reform", the conference planned more than 20 speech topics, such as performance optimization in the LLM era, best practices for accelerating the implementation of AI, and exploration of GenAI and general large model applications.
Shushi Technology was invited by the conference to appear in the large model exhibition area of the venue, and Cen Runzhe, general manager of financial digital products, brought a wonderful roadshow to show the newly released large model application product intelligent data analysis assistant swiftagent.
Booth situation: SwiftAgent, a digital potential intelligent analysis product, has attracted much attention
At the conference site, more than 10 large models and large models in the vertical field of digital potential have become the most beautiful scenery in this exhibition. In the exhibition area of Digital Potential Technology, you can not only see the core capabilities of SwiftAgent, a large-scale model application product of Digital Potential Technology, but also the application ideas and practices of products in the fields of big finance, pan-retail and high-tech manufacturing.
Dr. Li Fei, head of Digital Potential AI, introduced that Digital Potential Technology SwiftAgent was officially released last week, which is an innovative intelligent analysis product independently developed by Digital Potential Technology, which combines the most advanced large language model (LLM) and data analysis capabilities of Digital Potential Indicator Platform products, providing a powerful and intuitive tool for data analysts, business intelligence experts and non-technical users, without mastering query language or programming skills, to achieve natural language dialogue with data, mining data value and analyzing applications.
SwiftAgent product architecture diagram).
Roadshow: SwiftAgent supports applications in multiple industries such as pan-finance, high-tech manufacturing, and large retail
In the evening roadshow, Cen Runzhe, general manager of digital financial products of Digital Potential, delivered a wonderful speech "The Application of SwiftAgent in the Field of Enterprise Business Analysis", sharing the implementation experience and landing results of SwiftAgent in specific business scenarios, taking a financial institution project as an example, providing intuitive and practical product landing value for all experts present who are concerned about the application of large model technology, and stimulating the development of technical personnel and the hope of business personnel.
Cen Runzhe saidAt present, SwiftAgent has completed trials in some leading financial institutions to match their data analysis pain points. For example, before the launch of a leading brokerage, Datapotential found that its business team often encountered existing tool analysis dimension orders in the daily data analysis process.
First, the lack of flexibility and industry attributes are not obvious, and there are higher expectations for the intelligence and professionalism of data analysis tools.
Based on the technology of NL2MQ (Natural Language to Metric Query), SwiftAgent successfully combines the self-service data retrieval of the indicator platform with the semantic understanding and reasoning capabilities of the large model, and helps the brokerage realize scenarios such as natural language indicator data retrieval, analysis and attribution, and provides it with reports that fit the thinking of professional analysts, improves the depth of business analysis of the entire team, and improves the efficiency of data acquisition and use by dozens of times.
Digital potential swiftagent combines large model capabilities and indicator platform capabilities - application diagram).
As you can see,Digital potential swiftagentNo matter fromproductsThe core competence is still the industryLandingThe application has achieved initial resultsAt the same time, adhering to the attitude of open ecological cooperation, Datapotential cooperates with Zhipu AI and Baichuan Intelligence to provide customers with out-of-the-box integrated solutions for large models in the form of capability coupling, collaborative listing, and continuous iteration, and jointly coping with the challenges of large models in terms of long R&D cycle, high deployment threshold, high online cost, and business security.