Lead Scoring Model
AI-powered B2B lead scoring using LightGBM + SHAP for interpretability.
Usage
python3 score.py --input leads.csv --output scores.csv
Features
- CSV upload → scored leads with conversion probability
- Top features driving each score (SHAP)
- Ranked priority list
- Pipeline: LightGBM → SHAP → actionable insights
Input CSV Format
company_size,industry,page_views,email_opens,form_fills,job_title_score
50,tech,120,5,2,8
200,finance,45,2,0,5
Output
lead_id,score,probability,top_factor,risk_level
1,0.85,85%,page_views,hot
2,0.32,32%,low_engagement,cold
Notes
MIT-0 License | Requires: python3, lightgbm, shap, pandas