Craft Scoping Questions

Craft Scoping Questions

This documentation describes a workflow that generates scoping questions from a specification document. The workflow extracts relevant information to create questions that a bidder might ask, focusing on integration, user-related inquiries, and quality assurance. The workflow uses a Large Language Model (LLM) to not only generate the questions but also provide context and relevance for each.

Workflow Structure

The workflow is composed of several nodes, each focusing on a different aspect of the bidder question generation process.

1. RfP

   • Description: This is the input node where the RfP document is provided. It acts as the starting point for generating questions that a bidder might raise.

2. Integration Questions

   • Input: Receives the specification content.

   • Output: Extracts questions related to system integration. These questions are aimed at understanding how the solution will integrate with other systems, platforms, or software used by the client.

3. User-related Questions

   • Input: Processes information from the initial specification input.

   • Output: Generates questions related to user roles, permissions, and group-specific functionalities. These questions help clarify how the solution will handle user segmentation and authorization.

4. QA (Quality Assurance Questions)

   • Input: Receives data from the specification .

   • Output: Creates questions related to the quality assurance processes required by the client. This could include questions on testing procedures, performance criteria, and compliance requirements.

5. Contextualized Questions

   • Input: Combines output from the "User-related Questions" and "QA" nodes.

   • Output: This node synthesizes detailed bidder questions, including their relevance and context, based on the extracted RfP data. For example:

     • Question: "Could you describe the specific roles and permissions within the user group in detail?"

     • Relevance: "To ensure that the system meets the requirements of different user roles (e.g., event planners, administrators), this information is necessary for correct system configuration."

   • The output provides a complete question, followed by its relevance and the contextual information based on the input RfP document.

Workflow Control

The workflow is managed via a control panel on the right-hand side of the interface. The options available include:

• Run Workflow: Triggers the execution of the workflow, processing the specification and generating scoping questions.

• Open All Nodes: Expands all nodes, allowing users to view all inputs and outputs at a glance.

• Close All Nodes: Collapses the nodes for a more focused view.

Key Features

• LLM-Driven Question Generation: The workflow leverages a Large Language Model to intelligently extract and generate questions that are relevant to the RfP, providing a valuable tool for bidders.

• Contextualization: Each question generated is supported by contextual information that explains its relevance, making it easier to tailor proposals to client needs.

• Modular Design: The nodes allow for easy extension of the workflow, enabling the generation of additional questions depending on the complexity of the RfP or additional requirements.

Conclusion

This workflow simplifies the process of generating scoping questions by using AI to extract and contextualize information from a specification document. By focusing on integration, user management, and quality assurance, the workflow ensures that all critical aspects of the proposal are covered, helping requirements engineers to prepare detailed and well-informed questions that address the client's specific needs.