🚀 PromptDoc Challenge (June–July) 2026

🏆 This repository is my official submission for the Tips Hindawi Challenge (June–July) 2026.

👤 Participant

Field	Value
Full Name	Lojayn Khaled Farouk
Project Name	PromptDoc
GitHub Username	lojayn2004
Challenge Batch	June–July 2026
Training Program	Large Language Models (LLMs) Program
Organization	Edrak for Ai

---

📖 Project Overview

PromptDoc is an intelligent requirement analysis tool designed to bridge the gap between client discussions and technical specifications. It transforms raw voice/meeting transcripts into structured Product Requirement Documents (PRDs) and Gherkin-formatted user stories.

By leveraging Retrieval-Augmented Generation (RAG) powered by ChromaDB, LangChain, and Mistral AI, PromptDoc aligns the generated requirements with an organization's development guidelines, and documentation standards.

🏗️ AI Pipeline

---

✨ Features

• Voice/Text Transcription: Transcribes meeting audio files directly using OpenAI Whisper to convert client speech into text.
• RAG-Enriched PRD Generation: Retrieves context-relevant organization templates and standards from a ChromaDB vector store, injecting them into the LLM prompt to generate consistent technical specifications.
• Gherkin User Stories: Automatically outputs structured, behavior-driven development (BDD) user stories (Given-When-Then scenarios).
• Modern & Premium User Interface: A responsive and fluid React single-page application built with Vite, Tailwind CSS, and Framer Motion transitions.

---

🛠️ Technologies Used

Backend

• Python & FastAPI: Lightweight, high-performance web API framework.
• LangChain: LLM orchestration framework to build the RAG pipeline.
• ChromaDB: High-performance vector database for storing and retrieving document chunks.
• Sentence-Transformers: Open-source models for generating high-quality text embeddings.
• Mistral AI API: Large language model integration for generating specifications.
• PyMuPDF: Parsing and preprocessing PDF guideline documents.
• OpenAI Whisper: Local/API audio transcription.

Frontend

• React & Vite: Rapid, modern frontend development.
• TypeScript: Type-safe development environment.
• Tailwind CSS: Utility-first CSS styling.

---

⚙️ Installation

1. Clone the repository

git clone https://github.com/<your-username>/PromptDoc.git
git checkout main

2. Backend Setup

1. Navigate to the back directory:

   cd back

2. Create and activate a virtual environment:

   python -m venv venv
   # On Windows:
   venv\Scripts\activate
   # On macOS/Linux:
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

4. Create a .env file from the example:

   copy .env.example .env

5. Edit back\.env and fill in your API keys and directories:

   MISTRAL_API_KEY=your_mistral_api_key_here
   GROQ_API_KEY=your_groq_api_key_here
   EMBEDDING_MODEL=all-mpnet-base-v2

6. Run the FastAPI development server:

   uvicorn app.main:app --reload

3. Frontend Setup

1. Open a new terminal and navigate to the front directory:

   cd front

2. Install the frontend dependencies:

   npm install

3. Run the development server:

   npm run dev

---

🚀 Usage

1. Open your browser and navigate to the local frontend address (usually http://localhost:5173).
2. Upload a meeting audio recording or paste/type raw notes and transcripts in the input area.
3. Submit the transcript to run the RAG pipeline, which references internal templates to format the requirements.
4. View, copy, or export the generated Product Requirement Document (PRD) and structured Gherkin User Stories.

---

📸 Demo

https://drive.google.com/file/d/1NiyOxEfibdM0Hw9NSA4QXz-fKW5zljPe/view?usp=sharing

---

📈 Results

• Generates PRD & Context: Converts raw meeting transcripts to text and generates structured Product Requirement Documents (PRDs).
• React Preview: Previews the generated result dynamically in the React-based user interface.
• PDF Download: Allows downloading the final PRD output as a PDF file.

---

🔮 Future Improvements

• Try on Larger Data & Complex Cases: Test and optimize the model and RAG pipeline with larger datasets and more complex organizational scenarios.
• Multi-Format Export: Support exporting PRDs to PDF, DOCX, and Markdown formats.
• Direct Integration: Integration with project management tools like Jira or Linear to auto-create issues/tasks.

---

📚 About the Challenge

This project was developed as part of the Tips Hindawi Challenge (June–July) 2026.

Tips Hindawi is the internships department of Edrak for Ai, and the challenge encourages participants to build real-world projects, apply practical skills, and showcase their work through GitHub.

For more information about the challenge, training programs, and upcoming batches, visit the official Tips Hindawi website.

---

📄 License

This project is shared for educational and portfolio purposes.