README.md

metadata

license: cc-by-nc-4.0
task_categories:
  - text-classification
  - feature-extraction
  - tabular-classification
language:
  - en
  - multilingual
tags:
  - companies
  - business
  - lead-generation
  - b2b
  - firmographic
  - company-data
  - credit-scoring
  - financial-data
  - global-companies
pretty_name: World Company Database — Premium 1M (Revenue + Credit Score)
size_categories:
  - 100K<n<1M
configs:
  - config_name: default
    data_files:
      - split: train
        path: premium-1m-companies.parquet

World Company Database — Premium 1M Sample

1,000,000 curated companies with verified revenue data and credit scores, extracted from the S.C.A.L.A. Score global company database containing 272+ million records.

What Makes This Premium

Unlike random company samples, every record in this dataset has:

• Actual revenue data (revenue > 0) — no empty financial fields
• Credit score >= 50 — only creditworthy, financially assessed companies
• Sorted by revenue — the largest companies in each country come first

This is the top slice of 2.3 million financially enriched records out of 272M+ total.

Dataset Description

This dataset provides high-quality structured firmographic and financial data for 1 million companies across 13 European countries + the US, useful for:

• Financial analysis & benchmarking — Every record has real revenue, many have net income, assets, and equity
• Credit risk modeling — All records have S.C.A.L.A. credit scores (50-100) and letter grades
• Lead generation & B2B prospecting — Filter by country, sector, size, and financial health
• Market research — Analyze business landscapes with actual financial data
• ML training — High-quality labeled data for revenue prediction, credit scoring, sector classification

Schema

Column	Type	Description	Coverage
name	string	Company legal/trading name	100%
city	string	City / municipality	varies
country	string	ISO 3166-1 alpha-2 country code	100%
legal_form	string	Legal entity type (SAS, SA, SRL, etc.)	varies
sector	string	Industry sector code	varies
sector_desc	string	Sector description (human-readable)	varies
status	string	Company status (active, inactive, etc.)	varies
founded	string	Year or date of incorporation	varies
employees	integer	Number of employees	varies
revenue	bigint	Annual revenue (local currency)	100%
net_income	bigint	Net income (local currency)	varies
total_assets	bigint	Total assets	varies
equity	bigint	Shareholders' equity	varies
financial_year	integer	Year of financial data	varies
score	integer	S.C.A.L.A. credit score (50-100)	100%
grade	string	Credit grade (A/B/C/D/E/F)	varies
source	string	Data source identifier	100%

Note: tax_id and address fields are excluded from this public sample for privacy. Available via the Score API.

Country Distribution

Country	Records	Avg Score	Avg Revenue	Max Revenue
FR	400,000	53.5	18M	214B
NO	200,000	60.4	60M	941B
IT	200,000	77.0	20M	190B
PT	80,000	64.2	2.4M	29B
SE	50,000	65.0	17M	26B
BE	30,000	65.0	28M	92B
DK	27,000	65.1	270M	425B
CZ	6,000	70.0	1B	424B
EE	3,000	59.1	4.7M	2.2B
US	2,000	80.3	12.9B	717B
LV	1,000	76.3	82M	1.4B
ES	500	58.4	492M	62B
FI	500	54.2	169M	25B

Revenue values are in local currency (EUR for most countries, NOK for Norway, SEK for Sweden, CZK for Czech Republic, USD for the US, DKK for Denmark).

Usage

Python (pandas)

import pandas as pd
df = pd.read_parquet("hf://datasets/Alessandro114/world-company-database/premium-1m-companies.parquet")
print(df.shape)  # (1000000, 17)
print(df['revenue'].describe())

Python (DuckDB)

import duckdb

# Top 10 companies by revenue
duckdb.sql("""
    SELECT name, country, revenue, score, grade
    FROM 'hf://datasets/Alessandro114/world-company-database/premium-1m-companies.parquet'
    ORDER BY revenue DESC
    LIMIT 10
""").show()

# Country breakdown
duckdb.sql("""
    SELECT country, COUNT(*) as companies, AVG(score) as avg_score,
           AVG(revenue) as avg_revenue
    FROM 'hf://datasets/Alessandro114/world-company-database/premium-1m-companies.parquet'
    GROUP BY country ORDER BY companies DESC
""").show()

Python (datasets)

from datasets import load_dataset
ds = load_dataset("Alessandro114/world-company-database")

Full Database Access

This is a premium 1M sample from a database of 272+ million companies across 265 countries.

The full database contains 2.3M+ companies with financial data, and 272M+ total company records.

For full access with advanced filtering, enrichment, and real-time updates:

• Score API: https://score.get-scala.com/api — RESTful API with country, sector, revenue, and employee filters
• Bulk exports: Available for enterprise customers
• Custom enrichment: Tax ID validation, financial data, credit scoring

Built by S.C.A.L.A. — the enterprise AI operating system.

License

This dataset is released under CC BY-NC 4.0.

• Non-commercial use: Free with attribution
• Commercial use: Requires API access — see score.get-scala.com

Citation

@dataset{scala_score_premium_2026,
  title={World Company Database - Premium 1M Sample with Revenue and Credit Scores},
  author={S.C.A.L.A.},
  year={2026},
  url={https://huggingface.co/datasets/Alessandro114/world-company-database},
  license={CC BY-NC 4.0}
}