Connect a source and Basedash provisions a managed DuckDB warehouse that keeps every table synced, structured, and ready for agents to query. Already on Snowflake, BigQuery, or Databricks? Connect it instead.
14-day trial. No credit card required.
Powering analytics for 5000+ teams worldwide
Zero setup
There's no cluster to size and no database to create. Add your first connector and Basedash sets up the warehouse while you sign in, wires up the sync, and loads every source into its own schema.
Acme Warehouse
ReadyManaged DuckDB · created with your first connector
Schemas
salesforce14 tables
stripe21 tables
hubspotFirst sync
Built for agents
Agents answer better when every table lives in one place. Basedash joins Stripe, HubSpot, and your product data in a single DuckDB query, leaves out rows deleted upstream, and runs every question through your models and permissions.
Chat
What's net revenue retention by plan this quarter?
Searched 6 schemas · 118 tables
Used the governed NRR metric · Models
Joined stripe.subscription_history and hubspot.company · DuckDB · 0.8s
Enterprise leads at 118% NRR. Growth sits at 104% after two large expansions, and Starter is below 100%, so churn there outpaces upgrades.
NRR by plan · Q4
Target 100%
No lock-in
Your warehouse runs on DuckDB and standard SQL. Copy its credentials from the app and point any Postgres client, notebook, or script at the same tables Basedash queries, for the work you'd rather do outside it.
Basedash Warehouse credentials
Connect any Postgres client to the same tables Basedash queries.
- Postgres endpoint
- ••••••••••••••••.com:5432Copy
- Database
- acme_warehouseCopy
- Access token
- ••••••••••••••••3f9aCopy
Terminal
$ psql "$WAREHOUSE_URL"
acme_warehouse=> select count(*) from stripe.charge;
count
-------
48210
(1 row)
acme_warehouse=> Bring your own warehouse
Connect Snowflake, BigQuery, Databricks, Redshift, ClickHouse, or MotherDuck and Basedash queries it in place, with no copies to keep in sync. Sync SaaS tools into the managed warehouse next to it and use both in one workspace.
Data sources
SnowflakeDirect connectionQueried in place
BigQueryDirect connectionQueried in place- Basedash WarehouseManaged6 connectors
Operations and security
No database to administer and no capacity to plan. The controls your security team asks about come built in.
Fully managed
Provisioning, backups, scaling, and maintenance are on us.
Storage that scales
No storage cap to plan around as your data grows.
Point-in-time recovery
Daily backups you can restore to a point in time.
Encrypted everywhere
AES-256 at rest and TLS 1.3 in transit.
SOC 2 Type II
Audited annually. Customer data never trains models.
Hide data from AI
Keep schemas, tables, or columns out of AI context.
Access by group
Choose which groups and members can query each source.
Clean removal
A removed source's data is deleted after 30 days.
“We were looking for a way to get those kinds of queries more directly, and also aggregate different sources of truth — PostHog event data, database data, Stripe data — all in one centralized place.”
Jeffrey Zhao
Co-founder · Ordinal
Read case study →
“Even our engineers reach for Basedash instead of writing SQL. Answers that span multiple systems come back in minutes.”
Achraf Ghellach
Director of Engineering & AI · Matador AI
Read case study →
What Basedash Warehouse is, how it runs, and how it fits next to a warehouse you already have.
What is Basedash Warehouse?
Basedash Warehouse is a fully managed, cloud-hosted DuckDB warehouse that Basedash runs for you on MotherDuck. It's created automatically when you connect your first SaaS source, and every connector syncs into its own schema there. Basedash handles provisioning, backups, scaling, and maintenance, so your team can query Salesforce, Stripe, HubSpot, and the rest together in chat, dashboards, and the API without running warehouse infrastructure.
Why is the warehouse built on DuckDB?
DuckDB is a columnar engine designed for analytical queries, so aggregations and joins across large synced tables stay fast. It's also open source and speaks standard SQL, which keeps your data and queries portable. Nothing about the warehouse ties you to a proprietary format, and you can query the same tables from other tools with your warehouse credentials.
Can I query Basedash Warehouse from other tools?
Yes. Open the command menu (Cmd+K or Ctrl+K), search for Basedash Warehouse, and choose View credentials to get a Postgres-compatible endpoint, your database name, and an access token. Any Postgres client can connect with them, which is useful for notebooks, scripts, or cleanup you'd rather run outside Basedash. The option appears once at least one connector syncs into the warehouse.
Do I have to use Basedash Warehouse?
No. If your data already lives in Snowflake, BigQuery, Databricks, Redshift, ClickHouse, MotherDuck, or PostHog, connect it directly and Basedash queries it in place. The same goes for Postgres, MySQL, SQL Server, and other databases. Basedash Warehouse is where SaaS connectors land, so many teams use both: their own warehouse for core data and the managed one for tools like HubSpot or Stripe.
Is there a storage limit?
No. Storage in Basedash Warehouse scales with your data instead of being capped, and the warehouse is included in the Startup and Enterprise plans rather than billed as a separate product. The setting that matters more is sync frequency: frequent syncs keep fast-moving sources fresh, while daily syncs suit data that changes slowly.
What happens to warehouse data when I remove a source?
Removing a source marks it inactive and stops syncing new data right away. Its existing tables stay in the warehouse for 30 days so you can change your mind, and are then deleted automatically. If you need the data removed sooner, contact support and the team will delete it immediately.
How is data in Basedash Warehouse secured?
Warehouse data is encrypted at rest with AES-256 and in transit with TLS 1.3, and Basedash is SOC 2 Type II audited every year. Access runs through your workspace permissions: you choose which groups can query each source, hide sensitive schemas, tables, or columns from AI, and on Enterprise, review activity in audit logs. Customer data is never used to train AI models.