The streaming lakehouse for DuckDB

Turn streaming data
into answers with SQL

Ingest events, transform them with DuckDB SQL, and put them to work in analytics, search, and applications. Keep your history in open DuckLake storage. All in one system.

Talk to us Read the Quick Start
~1s
Hot commit interval
Configurable for your workload
1–10min
Parquet snapshot interval
Configurable for your workload
DuckDB
The SQL you already know
BoilStream extension

From incoming events to useful data

Connect your data streams. Transform, aggregate, and search with SQL. Bring the results to dashboards and applications, with history stored in DuckLake.

Your data streams flow into BoilStream, where DuckDB SQL transforms, aggregates, and searches the data. Recent data stays in DuckDB and history in DuckLake. Use the results in analytics, dashboards, and applications.
Send data. Work with SQL. Put the results to work. Explore the architecture
INSTALL boilstream FROM community;
LOAD boilstream;

-- Login with email, password, and MFA code
PRAGMA boilstream_login('https://your-server.com/user@example.com', 'password', '123456');

-- List catalogs, then attach the selected catalog_id
FROM boilstream_ducklakes();
PRAGMA boilstream_attach_ducklake('<catalog-uuid>', 'my_catalog');
USE my_catalog;
SELECT * FROM events;

One system. More ways to use your data.

Connect your producers, shape your data, and build the analytics and applications your team needs.

Use DuckDB SQL

Connect with the BoilStream extension and query your catalogs from DuckDB. Bring familiar SQL to event analytics, investigations, and application development.

Connect your producers

Send events using Kafka producers, Arrow, or SQL INSERT. Bring application logs, telemetry, and business events into the same analytics workflow.

Keep your data open

Retain analytical history in Parquet files managed by DuckLake. Keep your data in your object storage, with an open format built for analytics.

Work with your identity system

Connect Entra ID SSO and automated user provisioning. Manage users, roles, and catalog access through built-in administration tools.

Search with SQL

Add full-text search to selected fields and retrieve relevance-ranked results. Explore logs, documents, and events alongside your analytical queries.

Turn events into metrics

Build tumbling and sliding window aggregates with DuckDB SQL. Turn orders, sensor readings, and activity streams into useful summaries.

Transform data as it arrives

Filter, select, and enrich incoming rows with SQL. Create derived streams for the questions and applications that matter to your team.

Build dashboards and applications

Connect PostgreSQL-compatible tools or native DuckDB clients. Bring your data into operational dashboards, internal tools, and AI applications.

Control catalog access

Organize data in tenant-owned catalogs and share access with the right users. Manage permissions with catalog grants and role-based access controls.

Build more with the data already flowing

From operational analytics to rolling metrics and searchable knowledge, start with one SQL-native system.

Pricing

Same price whether you ingest 1 GB/day or 10 TB/day.
License tracks compute, not data.

Community

Self-serve evaluation
€0
Free, forever
8 vCPU · 2 pods
Cluster-wide totals
  • Full feature set (cluster mode, FlightSQL, hot tier, DuckLake 1.0)
  • Self-serve only
  • Docs & GitHub issues
  • Public Discord community
  • No staffed support, no SLA
Download
★ Back the project

Sustainer

Tip jar with priority Discord
€299
per year
8 vCPU · 2 pods
Cluster-wide totals
  • Same self-serve channels
  • Prioritized Discord channel
  • Funds the OSS-style work
Request invoice
Recommended

Starter

Smallest team running prod
€1,200
per year
16 vCPU · 4 pods
Cluster-wide totals
  • Email support, 3-business-day response
  • 2 onboarding hours included
  • All updates & releases
  • Suitable for production
Request invoice

Growth

Multi-pod cluster
€6,000
per year
64 vCPU / any pods
vCPU is the binding cap
  • Shared Slack channel
  • 1-business-day P1 response (EU hours) · P2/P3 best-effort
  • Quarterly check-in
  • 8 onboarding hours included
Request invoice

Scale

Enterprise · 24×7
Contact
Custom pricing
Any vCPU / pods
Cluster-wide totals
  • 24×7 support available
  • Named contacts
  • On-prem & architecture reviews
  • Custom support contract incl. SLAs
Contact
🇪🇺 EU e-invoicing — Finvoice / Peppol BIS
🔓 No vendor lock-in — Apache Parquet + DuckLake spec 1.0
🔄 Annual prepay, e-invoiced — Peppol within 1 business day; renew at the next tier when you grow

License expiry reverts the cluster to the free tier (8 vCPU / 2 pods). No data loss.

Tier change at renewal only. Never mid-year, never proration math.

Start with a streaming table

Deploy a single instance or scale horizontally with multiple nodes. Point to your S3 bucket and start ingesting.

Download BoilStream Contact Us