Page 46 - Touhpad Ai
P. 46

AI Project Framework
              For any task, we usually follow a plan, consciously or unconsciously. The sequence of steps taken from start to finish is
              called a Project Cycle, which ensures the task is carried out efficiently.
              Similarly, when developing an AI project, the AI project cycle provides a structured framework to guide us towards the
              desired outcome. It acts as a roadmap for planning, developing, and deploying AI solutions, offering a clear framework
              for organising, executing, and implementing the project.
              The following figure illustrates the six stages of the AI project cycle:

                                                              Problem
                                                              Scoping


                                                                                 Data
                                             Deployment
                                                                              Acquisition




                                                                                Data
                                             Evaluation
                                                                             Exploration



                                                              Modelling



               Stage 1   Problem Scoping
              To address a problem effectively, it is essential to begin by thoroughly understanding it. This phase focuses on analysing
              the various factors influencing the issue and determining how AI technology can offer a suitable solution.
              To gain a comprehensive understanding, this stage uses the 4W’s: Who, What, Where, and Why. This method ensures
              that all critical aspects of the problem are clearly identified, aligning the perspectives of stakeholders and team members.
              u Who: Identifies the individuals affected by the AI solution, including stakeholders, target users, and decision-makers
                 involved in the project.

              u What: Defines the specific problem or challenge that the AI project aims to solve, including its goals and desired
                 outcomes.
              u Where: Specifies where the AI solution will be implemented, whether in a technical environment, a particular domain,
                 or a geographical location.
              u Why: Explains the purpose of solving the problem, highlighting the value and impact for the business, users, or society.
              These questions provide a structured approach to ensure all relevant aspects are considered, facilitating more effective
              analysis and solution development. Once the 4W’s problem canvas is completed, the next step is to consolidate the
              information into a comprehensive problem statement template. This template compiles all key details in a single format,
              offering a clear reference for future use. A well-prepared problem statement ensures that the issue is clearly defined and
              understood by all stakeholders.

                                                   4W Problem Statement Template
                                Our                         [stakeholders]                    Who

                                has a problem that          [issue, problem, need]            What
                                when/while                  [context, situation]              Where
                                An ideal solution would     [benefit of solution for them]    Why


                 44     Touchpad Artificial Intelligence - XI
   41   42   43   44   45   46   47   48   49   50   51