Exploration is how you build intuition about data. But with Iceberg, exploration has traditionally required the same heavy infrastructure as production queries. PondPilot changes that — connect to an Iceberg REST Catalog from your browser and start exploring immediately.
Exploration-First Design
PondPilot is built for the discovery phase:
- Schema browser — see what columns and types exist before writing a query
- Quick preview — view sample rows without writing SQL
- SQL editor — write and iterate on queries with autocomplete
- Result export — save interesting subsets as CSV or Parquet for deeper analysis
Connect to Any REST Catalog
If your Iceberg catalog exposes a REST API (most modern catalogs do — Tabular, Nessie, AWS Glue via REST, Polaris), PondPilot can connect to it. DuckDB’s Iceberg extension in WASM handles the catalog protocol, metadata resolution, and data file reading.
Exploration Patterns
Data profiling:
SELECT column_name, COUNT(*), COUNT(DISTINCT column_name),
MIN(column_name), MAX(column_name)
FROM catalog.schema.table
Freshness check:
SELECT MAX(updated_at) as latest_record FROM catalog.schema.orders
Schema comparison — open two tables side by side and compare their structures.
Zero-Cost Exploration
There’s no compute cost to starting an exploration session. No Spark cluster billing by the minute. No Trino coordinator to keep warm. Your browser is the compute engine, and it’s already running.
For data teams that pay per-query or per-compute-hour, this means exploration doesn’t have to compete with production workloads for budget.
When to Use This vs. Production Infrastructure
Use PondPilot for: schema discovery, data previews, quick counts, validating recent writes, demoing data to stakeholders.
Use your production engine for: full table scans, complex joins across large tables, latency-sensitive dashboards.
Get Started
Open app.pondpilot.io and explore your Iceberg data.