Data Workshop & SQL Analytics Studio

Modern analytical datasets often originate in disparate formats—such as transactional CSV exports, nested JSON event payloads, and column-oriented Apache Parquet files. Traditional analysis typically requires spinning up remote databases or uploading gigabytes of sensitive proprietary data to third-party cloud analytics platforms. The Local SQL Data Workshop runs an enterprise-grade SQL OLAP database engine directly inside your browser sandbox using DuckDB-WASM and Web Workers.

You can load multiple independent tabular files simultaneously into ephemeral browser memory (:memory:), inspect schema column types and null counts, join tables across different file formats using standard ANSI/DuckDB SQL syntax, and execute analytical aggregations (GROUP BY, window functions, CTEs) in milliseconds. Advanced CSV ingestion options preserve leading zeroes (e.g. employee IDs and phone numbers), financial decimal strings (e.g. currency amounts), 64-bit integer precision, and ISO-8601 UTC timestamps without floating-point IEEE 754 precision loss. Large query result sets are displayed in a high-performance virtualized grid with dedicated controls to cancel long-running queries.

All database indexing, SQL execution, and table transformations occur 100% locally on your device. The analytical engine operates with external network access disabled; no file contents, schema names, or query strings ever leave your computer. Clean analytical outputs can be safely exported to Apache Parquet, CSV (with automated formula injection defense against spreadsheet macro attacks), or JSON.