🚀 RapidSegment – Strategic Segmentation & Scorecard Engine

RapidSegment is an industrial‑grade, combinatorial heuristic engine for discovering high‑lift predictive segments and compiling them into transparent, production‑ready scorecards. It bridges the gap between black‑box ML and legacy SQL rules engines.

Rapid Segment Interactive Explainer (Must Try)

Explainer

📘 Full documentation, architecture diagrams, statistical background, configuration reference, and FAQs live in the repo-level README. This page only covers install + a quick usage example.

📦 Installation

pip install rapidsegment

With the optional no-code Web UI:

pip install "rapidsegment[ui]"
rapidsegment-ui          # opens http://localhost:8501

⚡ Quick Start

import numpy as np
import pandas as pd
import duckdb
from rapidsegment import StrategicSegmentBuilder, StrategicSegmentScore
 
# 1. Synthetic data (or use your own)
np.random.seed(42)
n = 50_000
data = pd.DataFrame({
    "cust_id": [f"CUST_{i:05d}" for i in range(n)],
    "max_dpd_12m": np.random.choice([0, 15, 30, 60, 90], n, p=[0.7, 0.15, 0.08, 0.05, 0.02]),
    "utilization_avg_3m": np.random.uniform(0, 1.2, n),
    "risk_segment": np.random.choice(["Low", "Medium", "High"], n, p=[0.6, 0.3, 0.1]),
    "default_flag": np.random.choice([0, 1], n, p=[0.95, 0.05]),
})
 
# 2. Configure and run the builder
builder = StrategicSegmentBuilder(
    target="default_flag",
    top_n_vars=15,
    max_segments=5,
    ignore_features=["cust_id"],
)
segments = builder.extract_segments(data)
print(pd.DataFrame(segments)[["segment_id", "count", "lift", "sql_filter"]])
 
# 3. Build a scorecard
segment_cols = []
scoring_df = data[["cust_id", "default_flag"]].copy()
for seg in segments:
    col = f"SEG_{seg['segment_id']}"
    scoring_df[col] = duckdb.sql(f"SELECT ({seg['sql_filter']}) FROM data").df().astype(int)
    segment_cols.append(col)
 
scorer = StrategicSegmentScore(target_col="default_flag", primary_key="cust_id", segment_cols=segment_cols)
model = scorer.calculate_and_export_weights(scoring_df, "model.json")
print("Deciles:", model["decile_min_thresholds"])

🧩 How It Fits Together

Raw Data → StrategicSegmentBuilder → Segments (SQL rules) → StrategicSegmentScore → JSON Scorecard

StrategicSegmentBuilder finds high‑lift, hierarchical rules on your data; StrategicSegmentScore turns those rules into a weighted, deployable scorecard. UniversalDataLoader and BigQueryFeatureSelector are available for ingestion and feature screening — see the full README for details on every component, the extraction algorithm, statistical foundations, and the complete configuration reference.

📄 License

MIT — see LICENSE.

Built with ❤️ by Bishwarup Biswas. Special thanks to Guillermo Navas Palencia for Optbinning.

👉 For everything else — features, architecture, config reference, FAQs — see the repo README.