RPA执行能力
Search documents
科大讯飞“招采智能体平台”发布,评标效率提升70%
Xin Lang Cai Jing· 2026-01-16 08:07
Core Viewpoint - The launch of the "Procurement Intelligent Agent Platform" by iFLYTEK integrates large model decision-making and RPA execution capabilities, allowing companies to optimize procurement processes and reduce operational costs significantly [1][5]. Group 1: Platform Features and Capabilities - The platform can compress the average time for preparing bidding documents from 5-7 working days to just 30 minutes, with a bid detection accuracy rate of 96% and an overall evaluation efficiency improvement of over 70% [3][7]. - iFLYTEK's self-developed "Xingchen Agent" technology allows enterprises to assemble AI capability components with low-code or no-code methods, enabling quick customization of intelligent agents tailored to their business processes [3][7]. - The platform supports various scenarios such as "intelligent document editing," "anomaly behavior detection," and "assisted evaluation," enhancing the procurement process's efficiency and compliance [3][4]. Group 2: Impact on Procurement Efficiency - The intelligent evaluation system, co-built with the State Energy Group, has successfully operated over 180,000 procurement projects with an evaluation accuracy exceeding 97%, transforming the evaluation process from "1-2 projects per day" to "10 minutes for intelligent evaluation results" [3][7]. - In scenarios with over 1,000 pages of bidding documents and more than 100 bidders, traditional manual evaluations showed a scoring deviation rate of about 32%, while the platform achieved a 97% accuracy rate in objective evaluations and a 90% consistency rate in subjective evaluations [9]. Group 3: Industry Transformation - The platform represents a significant shift in the business model from project-based delivery to a scalable infrastructure that lowers application barriers and trial costs for enterprises, facilitating the widespread adoption of intelligent procurement technologies [4][8]. - The platform's multi-model collaborative mechanism for detecting bidding anomalies, including text and image similarity analysis, achieves a comprehensive recall rate of 96%, providing traceable and auditable compliance assurance for the procurement process [4][8].