Schema-first generation

Generate synthetic data from schemas

Great Generator is built for teams that already know their columns and need safe non-production data quickly.

Great Generator creates synthetic data. It does not anonymize, mask, de-identify, or transform production records.

Schema string

A compact schema string is useful in notebooks, examples, and quick tests.

from great_generator import generate_from_schema

df = generate_from_schema(
    schema="id int, customer_name string, email string, balance decimal(12,2)",
    rows=1000,
)

Semantic fields

Name-like fields, email fields, status fields, dates, amounts, IDs, and other common enterprise fields are detected from column names, not only raw types.

Output control

The function returns a DataFrame by default so users can write to CSV, Parquet, Delta, database tables, or cloud paths with their normal Pandas or Spark APIs.

df.to_csv("customers.csv", index=False)
df.to_parquet("customers.parquet", index=False)

Keep exploring

Getting started with Great GeneratorInstall Great Generator and create synthetic non-production data from schemas, SQL DDL, and related-table definitions for development and testing. SQL DDL synthetic data workflowsParse SQL CREATE TABLE statements into schema contracts and generate synthetic data for ANSI, Spark, and Databricks-style table definitions. Query-aware synthetic dataGenerate synthetic data with required values, partition coverage, selectivity targets, and join coverage for SQL and data engineering tests. API reference overviewReview the main Great Generator public APIs for schema generation, SQL DDL parsing, relational generation, CDC, anomalies, advisors, and exports. GitHub repositorySource code, issues, releases, and project history. PyPI packageInstall Great Generator from Python Package Index.