Your Parquet files. Fully analyzed. No servers. No code required.

Write SQL with DuckDB, visualize column distributions, detect outliers, run correlation heatmaps, and export to CSV, JSON, Excel or Parquet — all inside your browser. Business analysts and data engineers, both at home.

→ Open ParquetLab See features

clinical_data.parquet · 8,432,019 rows

Results Charts Analysis

SELECT study_code, COUNT(*) AS subjects, AVG(biomarker) AS avg_biomarker, SUM(adverse_event::INT) AS adverse_events
FROM clinical_data
WHERE visit_date >= '2024-01-01'
GROUP BY study_code ORDER BY subjects DESC

study_code ↑	subjects	avg_biomarker	adverse_events
ASTRID-2024	4,821	0.7823	143
BEACON-Q1	2,109	0.8341	87
DELTA-PILOT	1,502	0.6912	61

// capabilities

Not just a viewer.
A full data workspace.

01
✏️SQL Editor with DuckDB
Write any SQL against your Parquet files. Column and table autocomplete as you type. Format, history, keyboard shortcuts. Full DuckDB power in your browser.

        DuckDB WASM
        Autocomplete
        50-query history
      

02
🔬Advanced Analysis
Go beyond queries. Run a Pearson correlation heatmap, detect outliers with the IQR method, or aggregate time series by day, week, month, quarter, or year.

        Correlation
        Outliers
        Time series
      

03
📊Charts & Visualizations
Bar, Line, Area, Scatter, Pie, Donut. Split series by column, toggle labels, sort order. Filters apply live. Save as PNG. Sidebar controls always visible.

        6 chart types
        Live filters
        Export PNG
      

🗂️

Schema & Column Stats

Left panel lists all tables and columns. Click any column header to see stats: nulls, distinct values, distribution, correlations — all from your current query results.

⚙️

Filter, Sort & Manage Columns

Per-column filters, sort ascending/descending, show/hide columns, drag headers to reorder. Reset all filters at once.

💾

Export in 4 Formats

Download full results or filtered subsets as CSV, JSON, XLSX, or Parquet. Exports respect active filters — you get exactly what you see.

CSV JSON Excel Parquet

🔗

Multi-file JOIN

Drop multiple Parquet files to load them as separate tables. JOIN across them in SQL. Each file gets its own table name, derived from the filename.

☁️

Cloud Connect

Connect GCS via Google OAuth or HMAC keys. Connect Azure via Microsoft OAuth or SAS token. Your files stream directly to your browser. We handle auth — your data never touches our servers.

🔒

Zero Upload. Always.

DuckDB runs entirely in WebAssembly in your browser. No file ever reaches a server. Close the tab and everything is gone. File size limited only by your browser memory.

// who it's for

Built for everyone
who works with data.

📈

Business Analysts & Researchers

No Python, no Spark, no setup. Open your Parquet file, filter columns, explore distributions, and export to Excel in minutes. Parquet files have never been this accessible.

⚙️

Data Engineers & Analysts

Full DuckDB SQL, multi-file JOINs, schema inspection, correlation heatmaps, outlier detection. A fast, private scratchpad for any Parquet file — local or in the cloud.

🧬

Pharma & Clinical Research

PHI and PII never leave your browser. Auditable architecture, zero data retention. Safe for regulated environments where data cannot touch external servers.

☁️

Cloud Data Teams

Connect directly to GCS or Azure Blob with OAuth. Browse buckets, query files without downloading them. No keys to manage, no infrastructure to spin up.

// how it works

Open. Query.
Export.

Open your file

Drop a local Parquet, connect to GCS or Azure, paste a public URL, or try the built-in sample dataset. Schema and tables load instantly.

Query & analyze

Write SQL with full DuckDB support. Visualize with charts. Run correlation heatmaps, detect outliers, aggregate time series — all in the browser.

Export your results

Download filtered results as CSV, JSON, Excel, or Parquet. Share insights with your team in the format they actually use.

// cloud integrations

Connect to your
cloud storage.

Coming soon 🦆

Google Cloud Storage

Browse buckets and stream Parquet files directly to your browser. Files never pass through our servers.

Google OAuth HMAC keys

Coming soon ☁️

Azure Blob Storage

Access Azure Blob containers and query Parquet files without any data touching our servers.

Microsoft OAuth SAS token

Coming soon 🪣

Amazon S3

AWS integration coming soon. Same zero-upload guarantee — files stream cloud → browser directly.

AWS credentials

// privacy by design

Your data never
leaves your browser.

DuckDB runs entirely in WebAssembly inside your browser tab. No file, query, or result ever reaches a server. Close the tab and everything is gone.

Safe for PHI, PII, trade secrets, and any regulated data. Open the Network tab and verify it yourself — no surprises.

✓ DuckDB runs locally (WebAssembly) LOCAL

✓ Cloud files stream directly to your browser DIRECT

✓ Exports generated in your browser LOCAL

✗ Your data → ParquetLab servers NEVER

✗ Data stored, logged, or retained NEVER

Your Parquet files. Fully analyzed. No servers. No code required.

Not just a viewer.A full data workspace.

Built for everyonewho works with data.

Open. Query.Export.

Connect to yourcloud storage.

Your data neverleaves your browser.

Start analyzingright now.

Not just a viewer.
A full data workspace.

Built for everyone
who works with data.

Open. Query.
Export.

Connect to your
cloud storage.

Your data never
leaves your browser.

Start analyzing
right now.