AI-Powered Document Test Case Generator

I need a streamlined web-based workflow where I can upload spreadsheets, PDF process maps or Word specification documents and have an AI engine do two things for me: first, accurately extract the key information inside each file, and second, use those insights to automatically build a fully-formed suite of test cases. The finished cases must be delivered in a single Excel workbook, ready for immediate use.

Here’s the picture I have in mind:
• A simple UI (drag-and-drop or standard upload) that accepts the three document types mentioned above and shows real-time processing status.
• A backend that handles parsing (Pandas, pdfplumber, python-docx or similar), feeds the cleaned content to an LLM via OpenAI, Azure OpenAI, or a fine-tuned open-source model, and then formats the AI’s response into an .xlsx file with columns such as Test ID, Preconditions, Steps, Expected Result and Priority.
• Configuration files or prompt templates so I can adjust how the model phrases the test steps later on.
• Clear, well-commented code, a concise README and a handful of unit tests so I can maintain or extend the project easily.

Acceptance criteria:
1. When I upload one sample of each supported file type, the system returns an Excel workbook containing at least ten coherent, non-duplicate test cases.
2. All mandatory columns in the sheet are populated—no critical blanks.
3. The entire process runs unattended once files are submitted.

A proof of concept inside two weeks would be perfect, with time afterward for polishing and hand-off.

The AI should provide test cases able to be used for UAT with steps and navigation in the major ERPs, workday, SAP and oracle

Deadline: 4 Weeks POC

- I would need the ability enabled for licensing through the site/application
- Output should be able to be created in format ready to upload into the main test management tools