TANSUN(SZ:300872)2025-02-12 17:17Summary of Conference Call Company and Industry - The conference call primarily discusses the advancements and strategies of Tianya Technology in the field of AI and large models, particularly focusing on their collaboration with DeepSeek and the application of these technologies in the banking sector. Key Points and Arguments 1. Progress in AI and Large Models: Tianya Technology has made significant advancements in AI, particularly in large model development, which is seen as a core area for future growth and efficiency improvements in various applications, including marketing and intelligent recruitment [1][2][3]. 2. Strategic Partnerships: The company has signed a strategic cooperation agreement with Hunan University, committing to invest 200 million over five years to enhance their capabilities in supercomputing and large model research [3][4]. 3. Application in Marketing and Recruitment: The company is exploring applications of large models in marketing, such as deep tagging and intelligent merchant classification, as well as in recruitment processes to streamline resume screening [3][4][5]. 4. Development of Multi-modal Models: Tianya Technology is developing multi-modal large models tailored for different banking clients, addressing various needs such as marketing, risk assessment, and code generation [4][5]. 5. Support for Small and Medium Banks: The company aims to provide integrated solutions for small and medium-sized banks, which lack the resources of larger banks, by offering hardware, platforms, and products in a unified package [5][6]. 6. Training and Development: A training program has been established to help banking staff understand and utilize large models effectively, fostering a culture of collaboration between business and technology departments [6][7]. 7. Establishment of Intelligent Financial Laboratory: An intelligent financial laboratory has been set up to oversee the development of large models across various business units, ensuring alignment with organizational goals [7][8]. 8. Deployment of DeepSeek Platform: The company has successfully deployed its solutions on the DeepSeek platform, enhancing its capabilities in various financial applications [9][10]. 9. Data Dependency and Local Deployment: The development of large models is heavily reliant on data, and there are challenges related to local deployment due to data privacy concerns in the banking sector [11][12]. 10. Cost and Technical Considerations: The deployment of different versions of models (e.g., full and distilled versions) presents varying costs and technical challenges, which need to be addressed for effective implementation [13][14]. 11. Exploration of New Business Models: The company is considering new business models, such as SaaS and usage-based pricing, to provide AI solutions to banks, particularly in areas where reliance on internal data is minimal [15][16]. 12. Market Response and Adoption: There is a cautious but growing interest among banks, especially smaller institutions, in adopting large models following the release of DeepSeek, with increased urgency in procurement processes [26][27]. 13. Future Plans and Investments: The company plans to continue investing in AI and large model development, focusing on enhancing its engineering capabilities and collaborating with academic institutions to secure talent [35][36]. Other Important but Overlooked Content - The company emphasizes the importance of integrating AI into existing banking processes rather than creating entirely new systems, which could lead to incremental improvements in efficiency and service delivery [20][30]. - There is a recognition of the unique challenges faced by the domestic banking sector compared to international counterparts, particularly regarding the complexity of existing credit systems and regulatory environments [16][18]. - The call highlights the need for ongoing collaboration with leading banks to refine and adapt AI solutions to meet specific industry needs, ensuring that technological advancements align with practical applications [22][21].