I spend my days running deep-dive research on digital public infrastructure across all 54 African countries. Perplexity Max already gives me strong results when I supply well-structured metadata—research parameters, keyword lists, data-formatting notes and clear collection guidelines—but at the moment I must paste each country-specific prompt by hand, check every response, then copy the findings into my databases. It is time-consuming and error-prone, so I want to streamline the entire loop through a Windows command-line interface that talks directly to Perplexity’s API. The tool should let me: • feed a single template prompt and automatically batch it out for every country in my list; • monitor the responses in real time and flag or retry when Perplexity drifts off topic; • export the final answers in both CSV and Markdown so they drop straight into my analysis pipeline. I am an information architect, so I will provide the prompt templates, the exact metadata fields, and the naming conventions. Your job is to wrap them in a robust, well-documented CLI (PowerShell or native executable—whichever you prefer) with clear installation steps and a README that shows example commands and expected output files. Acceptance criteria 1. Single command triggers batch processing of at least 54 prompts and returns consolidated CSV and Markdown files. 2. Automated checks catch missing fields or off-topic answers and log them for review. 3. Source code and documentation are delivered in a version-controlled repository I can pull locally. If this first iteration goes smoothly I expect to extend the project into more sophisticated data-science tasks, so clean, maintainable code is essential. And my ideal partner will have a working knowledge of perplexity and how it differs from and integrates with the mainstream LLMs.