# Four-Level Muse Opportunity Dashboard

## PURPOSE
Turns Greg Isenberg's four-level personal-agent framework — (1) mining, (2) owning the list, (3) owning the room, (4) building technology — into an interactive dashboard. Users pick a niche and city (night nurses in Montreal, plumbers in Phoenix, furniture in Austin) and see lead quality, a niche x city opportunity heatmap, and cumulative revenue as they climb the levels.

## WHY THIS MATTERS
The video stresses that each niche and each city has a different demand and cost structure, and that the goal is progression — "you can take one of these small businesses that you start using MetaMuse with and then turn it into something bigger". Visualization is how you compare dozens of niche-city pairs at a glance and decide where to go deep before committing real time and money.

## HOW TO BUILD
1. Load mined leads into a Pandas DataFrame (or generate a seeded sample).
2. Add sidebar filters for `niche`, `city` and success fee.
3. Render a Plotly trust-score histogram per source (groups vs Instagram vs Marketplace).
4. Build a Plotly heatmap of opportunity score across niche x city.
5. Add a line chart projecting revenue across the four levels.

```python
fig = px.imshow(
    matrix, color_continuous_scale="Viridis",
    labels=dict(color="opportunity"),
)
st.plotly_chart(fig, use_container_width=True)
```

## WHEN TO USE IN PRODUCTION
Use it as the ops console for a multi-city agency: track which niches fill fastest and where Muse mining yields the best trust scores. It also works as an investor-facing demo of the four-level thesis. Do NOT use it to make per-client pricing decisions, nor as a substitute for the actual market test — it visualizes a model, it does not validate demand.

## HOW IT HELPS STUDENTS
Students learn Streamlit app structure, Plotly chart selection and Pandas aggregation, while internalizing a framework for turning a raw agent output into a business decision.

## STARTER CODE
`starter/main.py` builds the sidebar, a seeded `sample_leads` generator, a niche x city heatmap and a four-level revenue projection, with `# TODO` markers for wiring a real Muse export and Airtable.
