TRS(300229)2024-01-18 11:04Group 1: Company Overview and Recent Developments - The company, Toris Information Technology Co., Ltd., held an investor relations activity on January 17, 2024, to discuss business development and future strategic planning [2] - The company has accumulated over 200 billion data assets from various industries, demonstrating a daily data acquisition capability of over 100 million [3] - The company has launched industry-specific large models in media, finance, government, and public opinion since June 2023, establishing partnerships with data exchanges in several cities [3] Group 2: Financial and Regulatory Updates - The company is in the process of a private placement of shares, which was accepted by the Shenzhen Stock Exchange on December 8, 2023, but the approval timeline remains uncertain [3] - The implementation of the "Interim Provisions on Accounting Treatment Related to Enterprise Data Resources" on January 1, 2024, will enhance the representation of data assets in financial statements [3] Group 3: Financial Technology Services - The company offers services in financial digital transformation, including intelligent risk control, smart regulation, and precise marketing, aimed at enhancing digital operational capabilities for financial institutions [4] - The financial large model covers various business scenarios, including intelligent risk control and automated business processing, improving efficiency in contract approval and compliance checks [4] Group 4: Digital Government Initiatives - The company has developed a comprehensive business system for digital government, providing solutions for government website integration, smart regulation, and data services [4] - The company serves 80% of central and state council institutions, 60% of provincial governments, and 50% of municipal governments, reflecting its extensive reach in the digital government sector [4] Group 5: Challenges in Large Model Implementation - Key challenges in the application of large models include content compliance, model security, data quality, business integration, and project cost-effectiveness [4] - High-quality data is crucial for the effectiveness of large models, with a preference for training on large-scale datasets with hundreds of billions of parameters [4]