Universal Analytics Dashboard

Streamlit app for uploading CSV or Excel files, inferring a dashboard mapping, filtering and lightly cleaning the data, visualizing two charts, and generating cached AI chart insights.

Final Files

• app.py: final deployable dashboard
• requirements.txt: Python dependencies
• README.md: setup, deployment notes, and learning journal
• project_instructions.md: assignment prompt

Features

• Upload csv, xlsx, xls, or xlsm
• AI-first dataset mapping with rule-based fallback
• Editable mapping for datetime, grouping, category, and metrics
• Sidebar filters based on the chosen chart fields
• Lightweight data prep:
  • drop null rows by selected columns
  • drop zero rows by selected numeric columns
  • impute missing values using mode, mean, median, zero-fill, or forward-fill
• KPI summary row
• Time-series chart with configurable:
  • time grain
  • chart style
  • aggregation method
• Bar chart with configurable top-category limit
• Quick insights panel
• AI chart analysis with caching and reusable history

Local Setup

1. Create and activate a virtual environment.
2. Install dependencies:

pip install -r requirements.txt

3. Add your API key locally in .env:

OPENAI_API_KEY=your_key_here

4. Run the app:

streamlit run app.py

Deployment

Deploy app.py as the entrypoint.

Do not commit .env or .streamlit/secrets.toml.

For deployment, set OPENAI_API_KEY in the host platform's secret manager.

Examples:

• Streamlit Community Cloud: add OPENAI_API_KEY in app Secrets
• Render / Railway / Hugging Face Spaces: add OPENAI_API_KEY as an environment variable

The app checks for OPENAI_API_KEY in this order:

1. Streamlit secrets
2. Environment variables loaded from .env

If your current key has ever been committed, shared, or exposed publicly, rotate it before deploying.