Fynd AI Intern — Take Home (Prepared Repo)

Structure

• prompts/ - prompt templates for Task 1
• task1_prompting.ipynb - Notebook for prompting experiments
• web_app/user_dashboard - Streamlit user app
• web_app/admin_dashboard - Streamlit admin app
• data/ - shared CSV data file (submissions.csv)
• report.md - short report (convert to PDF)

Quick start (local)

1. Create virtual environment and install requirements:

   pip install -r requirements.txt
   

2. Open task1_prompting.ipynb in Jupyter and replace the call_llm stub with your LLM client (OpenAI/Gemini/OpenRouter).
3. Run the notebook to evaluate prompts on a sample dataset (data/yelp_sample.csv).
4. Run the user dashboard:

   streamlit run web_app/user_dashboard/app_user.py
   

5. In another terminal run the admin dashboard:

   streamlit run web_app/admin_dashboard/app_admin.py
   

Deployment

• Use Streamlit Community Cloud, Hugging Face Spaces, Render, or Vercel (for frontend+backend).
• Put your API keys into environment variables or Streamlit secrets; do not commit them.