Research and implementation of AI driven intelligent accounting models

Abstract The rapid development of artificial intelligence technology has driven a transformation in the accounting field. This paper proposes an Intelligent Accounting Model and designs an intelligent accounting architecture that combines the DeepSeek Large Language Model with Activiti workflow technology. Focusing on artificial intelligence technology in vertical accounting scenarios centered on high-quality data, as well as key workflow technologies based on BPMN specifications, this study provides an in-depth discussion of the intelligent accounting model. It explores the technical implementation details of intelligent accounting services from six aspects: business operations with high real-time requirements, personalized management accounting services, the adaptive integration of complex business processes and accounting, localized AI deployment and data governance, accounting risk identification and prevention, and external data interface access. These discussions are illustrated through specific examples, such as team performance assessments in petrochemical enterprises and customized financial reimbursements, aiming to provide technical solutions for the practical implementation of intelligent accounting. Furthermore, the paper presents an implementation case of an integrated intelligent accounting platform tailored to the business operations of shipbuilding enterprises. Ultimately, the intelligent accounting model and implementation technologies, based on the combination of artificial intelligence and workflow technology, offer feasible solutions and technical methodologies for the transition of accounting towards intelligence.