Making Sense of Developing Artificial Intelligence-Based System in Software Development Life Cycle Manner and Addressing Risk Factors

When dealing with real business problems at a company, developing an artificial intelligence system no longer only relies on the success of compiling a program. The program should meet the efficacy requirements of the business. In particular, focusing on a machine learning model to reach 100% accuracy will encounter a complex and hardly used end-to-end application; thus, the project does not satisfy a functional minimum viable product (MVP). An intricate project also considers good risk management, facilitating software development performances like user interface design, interpretability, robustness, and development costs. In addition, we found that the presence of literatures apparently discusses a different sense of meaning related to this topic, i.e. the use of artificial intelligence to enhance the performance of software development process. For this reason, this paper tries to make sense the terms of developing an artificial intelligence-based system in software development life cycle and risk factors that could lead to project's success with the spirit of sense-making of Dervin's theory. To this end, compared to other models, our proposed model presents the foundation of AI-SDLC in the waterfall model concerning the risk factors. Experimental evidence supporting our analysis is also discussed.