AI应用落地进展调研-医药

Summary of AI Application Progress in the Pharmaceutical Industry Company and Industry Overview - The company is involved in the pharmaceutical industry and is focusing on the application of AI technologies to enhance its operations and efficiency. The company has been investing in AI since 2024 and plans to continue expanding its efforts in this area [1][2][4]. Key Points and Arguments AI Investment and Strategy - The company's AI investment is projected to increase from less than $3 million in 2024 to $6 million in 2025, and to reach $10 million in 2026, primarily focusing on software development [1][4]. - The strategic approach is termed "system plus AI," which involves collaboration with partners like Kingdee to integrate AI capabilities into existing systems [1][9]. Cloud Deployment and Data Security - The shift to private cloud deployment is driven by concerns over data and knowledge security, with plans for potential local deployment in the future [1][5]. - Current business systems are primarily localized, with AI capabilities interacting through a hybrid cloud model, reflecting the high data security and regulatory requirements of the pharmaceutical industry [1][6]. AI Applications and Development - Initial AI applications were tool-based, assisting clinical researchers with tasks such as article writing. The focus for 2026 will be on enhancing system and data-level AI capabilities, including intelligent judgment in bidding systems [1][10]. - The company is building specialized databases for training models in professional knowledge scenarios and creating an internal knowledge base platform for general scenarios [1][11]. Team Structure and Skills Development - An independent AI team was established in 2024, focusing on product management and the internalization of development capabilities, with plans for traditional IT staff to learn programming languages and model training techniques [3][14]. Investment Priorities - The company prioritizes AI applications in sales, followed by supply chain, finance, and human resources. Specific initiatives include AI-driven compliance checks and market analysis tools [3][20]. Supply Chain Management - The end-to-end supply chain project aims to unify data management from sales to production and procurement, enhancing visibility and efficiency [21][22]. Vendor Selection and Collaboration - The company seeks advice from top consulting firms when planning AI projects and prefers partners that can provide comprehensive solutions [23][24]. AI Training and Model Selection - The company utilizes a combination of vector databases and knowledge graphs to improve knowledge recall accuracy, with a focus on multi-modal data processing [3][16]. - Various AI models are evaluated based on their performance in specific business contexts, with a preference for models that demonstrate higher recall rates in clinical applications [30][31]. Additional Important Insights - The company is currently using Alibaba Cloud for GPU server rentals, primarily utilizing NVIDIA GPUs, and is considering domestic GPU options for future resilience [33]. - The development of a smart training system for sales personnel is underway, aimed at reducing the burden of manual training and ensuring certification before onboarding [24][25]. - The distinction between system-level AI and productivity tool AI is highlighted, with system-level AI requiring more integration with existing IT infrastructure [26]. This comprehensive overview captures the company's strategic direction, investment plans, and operational focus within the pharmaceutical industry regarding AI applications.