RunRight UAE — Analytics Platform

What's Included

File	Purpose
app.py	Streamlit analytics app (8 pages)
requirements.txt	Python dependencies
RunRight_Scorer.html	Standalone HTML scoring tool (no server needed)
RunRight_UAE_Survey_Raw.csv	Raw survey data (2,000 respondents)
RunRight_UAE_Survey_Encoded.csv	Encoded/feature-engineered data
enriched.csv	Scored & segmented full dataset
arm_rules.csv	61 association rules
precomputed.json	Pre-computed chart data
models/	Trained sklearn models (pkl files)

Deploy to Streamlit Cloud

1. Push all files to a GitHub repo
2. Go to share.streamlit.io → New app → Select repo
3. Set main file path: app.py
4. Deploy — all dependencies install automatically from requirements.txt

Run Locally

pip install -r requirements.txt
streamlit run app.py

HTML Scorer

Open RunRight_Scorer.html in any browser — no installation needed.

• Drag & drop a CSV of new customers
• Models run 100% in-browser (50-tree RF embedded as JSON, ~1.8MB)
• Download scored results as CSV

Model Performance

Model	Metric	Value
RF Classifier	AUC-ROC	0.594
RF Classifier	Accuracy	0.620
RF Regressor	R²	0.967
RF Regressor	MAE	AED 109
K-Means	Silhouette	0.078
ARM	Rules	61

Customer Tiers

Tier	Personas	Focus
Tier 1	Trail & Ultra Specialist, Serious Age-Grouper	Invest heavily — highest LTV
Tier 2	Wellness Professional, Social Community Runner	Scale — growth engine
Tier 3	Aspirational Beginner, Casual Lifestyle Runner	Freemium — nurture & upgrade