원문정보
초록
영어
An organization with sufficient data and well-defined AI processes is said to be better positioned to seize AI opportunities than its competitors, and is referred to as an AI-matured organization. Unlike traditional software systems, AI systems depend on large data sets, raising ethical and operational risks throughout the life cycle. With increasing demands for responsible AI, standardized life cycle processes such as those by IEEE and ISO were introduced. This study investigates the potential disconnection between standardized AI life cycle models established by IEEE and ISO and real-world practice. By pinpointing where and how actual practices diverge from standardized processes, this study clarifies AI life cycle operations, which is crucial for advancing theoretical development on the connection between AI process capabilities and organizational AI maturity.
목차
1. Introduction
2. Literature Review
2.1 Role of Data in AI
2.2 AI System Life Cycle Processes
3. Research Method
4. Narratives about AI Life Cycle Processes
4.1 Overall AI Life Cycle Processes
4.2 Agreement Process
4.3 Organizational Project Enabling Processes
4.4 Technical Processes
5. Research Findings and Conclusion
5.1 Data Analysis
5.2 Agile method
5.3 MVP &CICD, Automation, Iteration
5.4 Cross-functional user-cantered team for near continuous innovation
5.5 Factory Like Automation
6. Limitations and Future research
References