Applying a Requirements-Focused Agile Management Approach for Machine Learning-Enabled Systems

1. Tick all that apply. Completed PhD Incomplete PhD Completed Master’s Incomplete Master's Specialization Completed Bachelor's Incomplete Bachelor's 2. Evaluation of Agile RequirementsOriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. Dear participant, Thank you for spending part of your valuable time to complete this questionnaire. Objective: To evaluate the application of agile requirements-oriented management, supported by Lean R&D principles, in the context of developing MVPs in R&D&I projects aimed at building machine learning (ML)-enabled systems. This research follows a high academic standard and is conducted anonymously. We will not associate your email address with your responses. For more information/questions, contact: Lucas Romão: lr[email protected] Marcos Kalinowski: [email protected] Júlia Condé Araújo: [email protected] Marina Condé Araújo: [email protected] Pontifical Catholic University of Rio de Janeiro (PUC-Rio) * Indicates required question 1.1 - What is your academic background? (Please select all options that apply to your academic background.) * 1.2 - What is the field of study of your undergraduate degree? * 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 1/14 3. Mark only one oval. 1 project 2 projects 3 projects 4 projects 5 or more projects 4. Mark only one oval. Up to 1 year of experience in software projects Up to 2 years of experience in software projects Up to 3 years of experience in software projects Up to 4 years of experience in software projects Up to 5 years of experience in software projects More than 5 years of experience in software projects 5. Other: Tick all that apply. Data Scientist Developer Product Owner Scrum Master UX/UI Designer Business Owner Tech Lead 1.3 - What is your experience in software projects in terms of number of projects (including the projects you are involved in at ExACTa)? * 1.4 - What is your approximate experience in software projects in years (including the time on projects you work on at ExACTa)? * 1.5 - What role(s) do/did you play during these projects? (Please select all roles you have performed) * 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 2/14 This section consists of questions concerning the application of the RequirementsOriented Agile Management approach for machine learning-enabled systems. Approach 6. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 7. 2.1 - Was the approach adequate to support achieving the project’s objectives? * 2.2 - Comment on what motivated your responses about the approach in relation to the project’s objectives. 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 3/14 8. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 9. During the Initial Specification, the team specifies the machine learning components based on PerSpecML, promoting a shared understanding between the business, software, and machine learning teams. In this stage, requirements are elicited from the perspectives of system objectives, user experience, infrastructure, model, and data. * 2.3 - The Initial Specification is adequate to promote a shared vision among business, software, and machine learning teams about the objectives and requirements of the machine learning-enabled system. 2.4 - Comment on what motivated your responses about the Initial Specification. 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 4/14 10. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Concordo totalmente 11. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree During the Conception stage, the requirements are refined collaboratively. The Product Owner structures the Software and ML Backlogs with assistance from the software and ML teams, also defining the infrastructure and architecture Enablers that support the integration between the software and machine learning components. * 2.5 - The Conception stage is adequate for building the product backlog based on the requirements identified in the Initial Specification. 2.6 - The separation between the Software and ML Backlogs facilitates the organization of requirements and the integration of system and model responsibilities. * 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 5/14 12. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 13. 14. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Concordo parcialmente 5 - Concordo totalmente 2.7 - The use of infrastructure and architecture Enablers contributes to the integration between the software system and the ML pipelines. * 2.8 - Comment on what motivated your responses about the Conception. During the Technical Feasibility stage, the software and ML team assesses the technical feasibility of ML components and their integration with the system through Data Feasibility and Model Feasibility analyses and the creation of a Demo API (LoD 0). * 2.9 - The Technical Feasibility stage is adequate for anticipating and dealing with technical uncertainties in the agile management of machine learning-enabled systems. 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 6/14 15. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 16. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 17. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 2.10 - Data Feasibility is effective in ensuring the availability, quality and representativeness of the data necessary for the development of ML models. * 2.11 - Model Feasibility contributes to evaluating the suitability of algorithms and the complexity of the proposed models. * 2.12 - The Demo API (LoD 0) is useful for anticipating the integration between the ML model and the software system. * 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 7/14 18. 19. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 20. Mark only one oval. Não se aplica 1 - Discordo totalmente 2 - Discordo parcialmente 3 - Não concordo, nem discordo 4 - Concordo parcialmente 5 - Concordo totalmente 2.13 - Comment on what motivated your responses about Technical Feasibility. During the Development phase, system development is conducted iteratively and integratively, following Agile4MLS principles and maintaining synchrony between software and ML deliveries with the principles of Layers of Done (LoD) and Minimal Viable Model (MVM). * 2.14 - The use of joint iterations and sprints between software and ML teams contributes to the continuous delivery of value. 2.15 - O princípio de a equipe de ML trabalhar “duas sprints à frente” da equipe de software favorece o fluxo de desenvolvimento e integração. * 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 8/14 21. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 22. Mark only one oval. Not applicable 1 - Strongly disagree 2 - Partially disagree 3 - Neither agree nor disagree 4 - Partially agree 5 - Strongly agree 23. 2.16 - The use of the Layer of Done clearly communicates the level of maturity and reliability of ML models. * 2.17 - The definition of the Minimal Viable Model (MVM) is adequate for validating business hypotheses before the complete optimization of the models. * 2.18 - Comment on what motivated your response about the Development phase. 11/15/25, 9:01 PM Evaluation of Agile Requirements-Oriented Management in R&D&I Projects aimed at building Machine Learning-enabled Systems. https://docs.google.com/forms/d/1EvnlsjvtI6AIE3kngh8q6RWMJHbnwV64AXcwfmPyOFs/edit 9/14