Enhancing Requirement Gathering with LLMs

Requirement gathering meeting for a product

Well begun is half done. In the context of documentation life cycle management, it is the gathering requirements phase. It involves identifying the expectations of the audiences such as developers, end-users, or clients for a particular product or the documentation of the product to be precise. This process helps define what level deep to go when we document features, the scope, key functionalities, workflows, and any technical specifications that must be addressed.

Technical writers can effectively leverage large language models (LLMs) to enhance the process of gathering requirements. These models, whether open-source solutions like ChatGPT or proprietary internal LLMs, provide valuable support in streamlining communication and extracting key insights. The following steps outline how LLMs can be utilized for gathering requirements:

  • Automating Meeting Summaries
  • Processing Large Volumes of Input
  • Knowledge Base Search
  • Facilitating Stakeholder Communication
  • Generating Requirement Templates
  • Collaborative Brainstorming

Many LLMs also provide capability to attach the file which can include the initial information, script from the meeting so it is easier to process the text or information.

LLMs can streamline communication by summarizing complex discussions or interactions with stakeholders, generating structured questions or assisting in formulating requirement gathering questionnaires to ensure no critical details are overlooked.

Let’s assume that we are gathering requirements for Cloud-based Project Management Software. You can leverage LLMs to: 

  • Summarize key takeaways, action points, and critical requirements, ensuring that the documentation team captures all necessary details for the next stages of the documentation life cycle after requirement-gathering meetings.
    Example prompt: Summarize the key takeaways, action points, and critical requirements from today’s meeting about the cloud-based project management software. Ensure to include any specific requests related to task tracking, user permissions, and integration with third-party tools like Slack or Google Calendar.
  • During initial phases of requirement gathering, LLMs can participate in brainstorming sessions by suggesting potential use cases, edge cases, or functionalities that should be documented based on product features as product highlights.
    Example prompt: Suggest potential use cases, edge cases, or core functionalities that should be documented for the project management software. 
  • Define the objectives of your communication, such as clarifying product features, gathering user requirements, or understanding pain points. This helps LLMs generate targeted responses and relevant questions.
    Example prompt: Generate a set of objectives for stakeholder communication. Focus on clarifying the product’s task-tracking capabilities, gathering user preferences for dashboard customization, and identifying pain points related to project reporting.
  • Draft structured, open-ended questions that guide stakeholders toward sharing detailed requirements. You can prompt the LLM with product-specific details to ensure the questions are tailored.
    Example prompt: Draft a set of open-ended questions to guide stakeholders in providing detailed feedback on the project management software. Focus on questions related to feature expectations, workflow automation, and pain points in existing project management tools.
  • Use LLMs to summarize key takeaways after stakeholder meetings or interviews. This helps condense complex information into actionable insights, ensuring no important details are missed.
    Example prompt: Summarize the critical feedback and requirements gathered from the interview with the stakeholders about the project management software. Highlight their concerns regarding mobile accessibility, task prioritization, and reporting features.
  • At times, stakeholders provide feedback that is unclear, vague, or incomplete, it’s important to refine and clarify that information to ensure the documentation is accurate and meets the project’s requirements. LLMs can analyze the feedback and highlight where details are missing or ambiguous, allowing the writer to focus on areas needing clarification. Also, LLMs can assist by formulating specific follow-up questions to gain clearer information.
    Example prompt: Based on the feedback received as ‘better task assignment options’ specify which aspects of the interface are difficult to use? Are they referring to navigation, layout, or response time? 
  • LLMs can assist in creating templates or outlines for requirement documents based on best practices, which can be adapted to the specific product or project needs.
    Example prompt: Create a template for a requirement document based on best practices for project management software. Include sections for feature specifications, user roles, integration requirements, and security protocols, and so forth.
  • LLMs can help rephrase the feedback in more precise terms and suggest ways to communicate with stakeholders to ensure both parties are aligned.

Example prompt: Stakeholders feedback says, “The project dashboard feels cluttered and doesn’t provide enough information.” Rephrase the feedback to make it sound precise and clear in tone and provide a set of follow-up questions. 

In such and many other ways, LLMs can reduce the great amount of repetitive and time-consuming tasks and enhance communication efficiency, helping stakeholders articulate their needs in significantly less time. And they ensure that the requirements gathering is comprehensive and structured efficiently.

Need help perfecting your documentation process? Download our free checklist for gathering technical writing requirements or explore our comprehensive guide for mastering technical documentation.

Published by BestOptimizer

Writing and Analysis

Leave a Reply

Discover more from Technical Writing, AI Writing, Editing, Online help, API Documentation

Subscribe now to keep reading and get access to the full archive.

Continue reading