Interactive Prompt Engineering Assistant

This web application is a multi-purpose tool designed to assist in the prompt engineering process for credit analysis. It serves several functions:

1. A README: Explaining its purpose, functionality, and limitations.
2. A Prompt Engineering Guide: Showing how a detailed prompt for a Large Language Model (LLM) is constructed.
3. An Interactive Report Generation Tool: Allowing you to input company data and generate a simulated credit report.
4. An LLM-powered report generator: Allowing you to generate a report using a Generative AI model.
5. A Feedback and Evaluation Tool: Allowing you to score and provide feedback on the AI-generated report.

1.1. How it Works

The process is straightforward:

1. User Input: You provide company-specific data through the interactive form below. You can also upload a PDF or TXT file for additional context.
2. Prompt Construction: The tool takes your inputs and dynamically builds a detailed, structured prompt.
3. Report Generation: Clicking the generate button triggers two processes:
   • A simulated report is created instantly using internal, rule-based logic.
   • An AI-generated report is created by sending the constructed prompt to a large language model.
4. Output Review & Evaluation: You can review both reports, compare them, and provide feedback on the AI's performance to create a data record for future training.

1.2. Limitations

• API Key Required: To generate a report using an AI model, you must provide your own API key. The application does not store or transmit this key anywhere except to make the direct call to the model provider's API.
• Simulation Quality: The simulated report uses a simple, rule-based approach. Its output is a structured first draft and lacks the nuanced reasoning of a state-of-the-art LLM.
• Input Dependency: The quality of both reports heavily depends on the detail and accuracy of the information you provide.