An Interview with ChatGPT about Business Analysis
ChatGPT as a Business Analyst?
Since its introduction last November, ChatGPT has gained widespread attention and popularity as a highly intelligent and articulate chatbot. It converses with a person, assuming a wide variety of roles. For example, it can play the role of a business analyst interviewing stakeholders for business needs.
What impact will ChatGPT have on business analysis with its access to online knowledge and communication ability? It speaks for itself in the interview below.
ChatGPT Interview
PA: Introduce ChatGPT to business analysts in software development.
ChatGPT: Hello, business analysts in software development! I'm ChatGPT, a large language model trained by OpenAI. As a language model, I'm designed to process and understand human language and generate responses that are contextually relevant and useful.
As a business analyst, you may be involved in a variety of tasks such as requirements gathering, process improvement, data analysis, and project management. I can help you with many of these tasks by providing quick access to information, assisting with research, and even generating reports and summaries based on your inputs.
For example, if you're working on a project that involves a new software feature, I can help you by answering questions about the feature's requirements and capabilities, suggesting alternative solutions, and even providing examples of similar features in other software products.
Furthermore, I can assist you with data analysis tasks, such as exploring datasets, identifying patterns and trends, and generating visualizations. I can also help you with project management by providing status updates, creating task lists, and even coordinating with other team members.
Overall, as a business analyst in software development, I'm here to assist you in any way I can. Feel free to ask me any questions or provide me with any input, and I'll do my best to provide you with helpful and informative responses.
PA: A software project often starts with undefined or unknown requirements requiring discovery sessions with stakeholders. How could you help with that?
ChatGPT: As a language model, I can help you with the discovery sessions by providing assistance with the following tasks:
Facilitating communication: I can help you facilitate communication with stakeholders by providing language translation, suggesting communication strategies, and helping you to ask the right questions. This can help to ensure that all stakeholders are heard and that their needs and expectations are taken into account.
Providing context: I can provide context to the discovery process by drawing on my vast knowledge base to provide information on industry trends, best practices, and other relevant information. This can help you to identify potential requirements and features that may not have been considered otherwise.
Generating ideas: I can help you generate ideas for potential features and requirements based on the input provided by stakeholders. By using natural language processing and machine learning algorithms, I can analyze the language used by stakeholders to identify key themes and ideas that can be further developed into potential requirements.
Summarizing results: After the discovery sessions, I can assist you in summarizing the results by generating reports, summaries, and other documentation. This can help to ensure that all stakeholders have a clear understanding of the requirements and expectations for the project.
Overall, by leveraging my natural language processing and machine learning capabilities, I can assist you in the discovery process by facilitating communication, providing context, generating ideas, and summarizing results. This can help to ensure that the project requirements are well-defined and that all stakeholders have a clear understanding of the project goals and expectations.
PA: How can you provide context to the discovery process when it depends mostly on organization-specific knowledge that is not in your vast knowledge base?
ChatGPT: You are correct that organization-specific knowledge may not be available in my vast knowledge base. However, I can still provide some context to the discovery process by leveraging the information that is available to me.
Firstly, I can use my vast knowledge base to provide general industry context and best practices that may be applicable to your organization. This can help you to understand the broader context in which your organization operates, and identify potential requirements and features that are commonly used in the industry.
Secondly, I can use my ability to process natural language to analyze the language used by stakeholders during the discovery process. By doing so, I can identify key themes, concepts, and ideas that are specific to your organization. For example, I can identify commonly used terminology or acronyms, and use that information to better understand the organization's unique context.
Thirdly, I can learn from previous interactions with your organization. As a language model, I have the ability to store and recall previous interactions, which can help me to better understand your organization's specific context over time. This can help me to provide more accurate and relevant information as I learn more about your organization.
Finally, it's important to note that while I can provide some context to the discovery process, I am not a substitute for human expertise and knowledge. It's important to work closely with subject matter experts within your organization to ensure that all relevant knowledge is taken into account during the discovery process.
PA: Suppose a software project will replace an older solution. How would you learn about the older solution?
ChatGPT: To learn about the older solution that a software project is intended to replace, I would need access to information about the system, such as documentation, user manuals, or other resources that describe the system's features, functionalities, and requirements.
Here are some ways I could learn about the older solution:
Reviewing available documentation: If documentation is available for the older solution, I can analyze it to gain insights into the system's features, functionalities, and requirements. This can help me to understand how the system is currently being used, and identify areas for improvement or optimization.
Analyzing user feedback: If users of the older solution have provided feedback, I can analyze it to gain insights into their experiences with the system. This can help me to identify pain points or areas for improvement and inform the development of the new solution.
Conducting interviews: I can interview stakeholders who are familiar with the older solution, such as users, developers, or business analysts, to gain insights into their experiences with the system. This can help me to identify key features and functionalities of the system, as well as areas for improvement.
Analyzing system data: If data is available from the older solution, I can analyze it to gain insights into how the system is being used. This can help me to identify patterns or trends in usage, and inform the development of the new solution.
Overall, by leveraging these methods, I can learn about the older solution and gain insights into its features, functionalities, and requirements. This can help to inform the development of the new solution and ensure that it meets the needs and expectations of users and stakeholders.
PA: Once you have learned the organization and what it needs from a solution, what deliverables would you provide for a business analyst to evaluate and pass on to a software architect?
ChatGPT: Once I have learned about the organization and its requirements for a new software solution, I can provide the following deliverables to a business analyst for evaluation and pass them on to a software architect:
Requirements Document: A requirements document outlines the functional and non-functional requirements for the new software solution. It can include use cases, user stories, and acceptance criteria to help ensure that the new solution meets the needs of the organization.
Use Case Diagrams: Use case diagrams provide a visual representation of how users interact with the software solution. They can help to ensure that the solution meets the needs of the organization and its users.
User Interface (UI) Mockups: UI mockups provide a visual representation of the user interface for the new solution. They can help to ensure that the new solution is easy to use and meets the needs of users.
Data Model: A data model provides a visual representation of the data structures and relationships within the new solution. It can help to ensure that the solution is designed to handle the data requirements of the organization.
Technical Architecture Diagram: A technical architecture diagram outlines the high-level technical architecture of the new solution. It can include details on the technology stack, deployment architecture, and integration points with other systems.
Overall, these deliverables can provide the business analyst with a clear understanding of the requirements and expectations for the new solution, and can help the software architect design a solution that meets those requirements.
A Cocky Chatbot
ChatGPT expresses confidence in its natural language processing capabilities. It demonstrates good enough conversational skills in the interview, at least in a business context. It also gives us a decent sample of its business analysis knowledge.
ChatGPT asserts that it could help business analysis a lot with its communication skills. For example, a subject matter expert may prefer to interact with ChatGPT if it understands her domain. A business analyst could follow up on the conversation, asking ChatGPT to define and clarify domain-specific terms, thus relieving the subject matter expert of that task.
While ChatGPT has some understanding of a conversation and can point out inconsistencies, it also needs to identify communication gaps. For example:
Has a stakeholder expressed a business need sufficiently?
What alternative paths should a process take?
Does a use case require an undo function?
Business analysis will still need people to fill those gaps, as well as interview stakeholders who prefer a more human touch and create and present diagrams.
ChatGPT can boost business analysis productivity to the extent it knows about an organization, its domains, and its projects.
If ChatGPT has helped you with business analysis or solution architecture, please share your experience in the comments.