Self-host ParquetReader for private data workflows
Run SQL on Parquet, CSV, Avro, ORC, Feather, GeoJSON and NetCDF inside your own infrastructure. Built for teams that cannot upload sensitive data to hosted tools.
- Keep sensitive data and logs on your own systems. No uploads to ParquetReader Cloud and no hidden tracking.
- Query files directly from S3-compatible storage instead of copying them to a tool.
- Deploy as a Docker container or on Kubernetes.
Short demo
Request a self-hosted proposal
Tell us about your deployment setup and use case. We reply within 1 business day.
Why teams self-host
Built for environments where data cannot leave your infrastructure, while still enabling fast exploration and prototyping.
Data stays in your environment
Operate entirely within your environment for compliance, audits and internal policies.
Flexible Deployment
Docker, Compose or Kubernetes. Single node or scaled out.
Advanced SQL Querying
Fast filtering, joins and aggregations on file-based data without ETL.
ParquetReader API
Connect internal systems, dashboards and analytics tools through the API.
SQL directly on S3-compatible storage
Point ParquetReader at a bucket or prefix in S3-compatible storage and write SQL. It scans the relevant files and streams back results without moving data out of your account.
- Query Parquet, Feather, Avro, CSV, GeoJSON and ORC from S3-compatible storage.
- Column projection and pagination for efficient work with large files.
- Works with AWS S3, MinIO and custom S3-compatible endpoints.
Click to enlarge. SQL over multiple Parquet files in S3-compatible storage with instant preview and download.
Self-hosted ParquetReader is for teams that cannot send data to hosted tools and need to explore, validate and prototype on real files.
Share your deployment setup (VM/Kubernetes, S3/MinIO, auth requirements) and we will respond with a clear proposal.