Explore JSON Data Locally with SQL

API responses, log files, config exports — JSON is everywhere. But exploring nested JSON in a text editor is tedious. PondPilot lets you query JSON files with SQL, locally in your browser.

SQL on JSON, No Scripting

Forget writing Python scripts to parse JSON. DuckDB handles nested JSON natively:

SELECT
  json->>'name' as user_name,
  json->>'email' as email,
  json->'address'->>'city' as city
FROM events.json
WHERE json->>'status' = 'active';

PondPilot runs DuckDB in WebAssembly, so this all happens in your browser tab. Your JSON files stay on your machine.

Flatten Nested Structures

DuckDB’s UNNEST makes quick work of arrays and nested objects:

SELECT id, UNNEST(items) as item
FROM orders.json;

Turn deeply nested JSON into flat, queryable tables without writing a single line of transformation code.

Handle Large JSON Files

Browser-based doesn’t mean toy-sized. DuckDB WASM handles JSON files that would choke a web-based JSON viewer. Load multi-megabyte API dumps and query them efficiently.

Newline-Delimited JSON (NDJSON)

Working with log files or streaming data exports in NDJSON format? PondPilot handles those too. Each line is a row, ready for SQL.

Join JSON with Other Formats

Open a JSON file alongside a CSV and join them:

SELECT j.*, c.category_name
FROM products.json j
JOIN categories.csv c ON j->>'category_id' = CAST(c.id AS VARCHAR);

Export Clean Results

Once you’ve flattened and filtered your JSON data, export the results as CSV. Great for sharing with non-technical teammates who need the data in a spreadsheet.

Try It

Open PondPilot, drop a JSON file, and start querying. No signup, no upload, no backend.