While traditional intelligent archive systems still have issues like slow response times, they nonetheless play a significant role in developing talent for CBEC (Cross-border E-commerce). This article used big data technology to construct an intelligent archive system, improve the efficacy and caliber of CBEC talent nurturing, and more to better satisfy the actual needs of CBEC business development. The article first looked at the main structure of the system’s composition before summarizing the key ideas and methods for the hardware and software designs of the system; subsequently, the key technologies involved in building the system were discussed, and convolutional attention layers were introduced to explore in detail the methods of data extraction. Finally, to verify the actual performance of the intelligent archive system constructed using big data technology, it was compared with traditional systems. Experiments showed that when the number of archives reached 2100, the big data-driven system in this article had a response time of only 0.019 ms, while the traditional method’s response time was 0.141 ms. The system described in this article had a response time that was noticeably faster than that of traditional systems. Research has demonstrated big data technology’s usefulness in creating intelligent archive systems, providing more opportunities for promoting the efficiency of CBEC talent cultivation and the development of e-commerce enterprises.