# AI Adoption Maturity Roadmap Dashboard (Streamlit + Plotly)

## PURPOSE
A dashboard that visualizes the three-stage AI maturity path the Halliburton SVP lays out in the transcript: today AI is "primarily centered around personal productivity," the next shift is AI becoming "a co-worker in a team" that offers constructive advice, and beyond that comes "physical AI" for an industrial company. The dashboard scores each division's readiness (data house, domain context, agent capability) and places it on the roadmap, so leadership can see who is ready to move from assistant to co-worker and who still must fix the foundations.

## WHY THIS MATTERS
The video's central warning is that teams "dive in deep with AI at the beginning and then stall because your data isn't ready" and "your domain context isn't quite correct." A maturity view makes that failure mode visible before it happens: a division at the assistant stage with weak data readiness should not be pushed to co-worker or physical AI. It turns a strategic narrative into a measurable portfolio.

## HOW TO BUILD
1. Define a deterministic dataset of divisions with readiness scores (0-1) for data, context, and agent capability.
2. Compute a stage label: `assistant`, `co-worker`, or `physical AI`.
3. Use Streamlit sidebar filters (division, minimum readiness) to drive the view.
4. Render a Plotly scatter (data readiness vs. context readiness) with stage as color, plus a radar chart per division.
5. Add a summary table and a callout for divisions flagged "stall risk."

Key snippet from `starter/main.py`:
```python
def stage_for(data: float, context: float, agent: float) -> str:
    if data >= 0.8 and context >= 0.8 and agent >= 0.7:
        return "physical AI"
    if data >= 0.6 and context >= 0.6 and agent >= 0.4:
        return "co-worker"
    return "assistant"

df["stage"] = [stage_for(d, c, a) for d, c, a in
               zip(df["data_readiness"], df["context_readiness"], df["agent_capability"])]
```

## WHEN TO USE IN PRODUCTION
Use it in quarterly AI portfolio reviews, executive briefings, or as a steering-committee dashboard tracking progress against a roadmap. Do NOT use it for real-time operations monitoring or as an authoritative source of truth — it is a strategic planning view fed by periodic assessment, not live telemetry.

## HOW IT HELPS STUDENTS
Students learn Streamlit layout and widget patterns, Plotly express/scatter visualization, and how to encode a qualitative maturity model as reproducible, filterable code. Building a dashboard that non-engineers can navigate is a highly transferable skill.

## STARTER CODE
See `starter/main.py` — a Streamlit + Plotly skeleton with deterministic readiness data, stage classification, and `# TODO` markers for loading real assessments and adding radar charts.
