New: give any agent a live, governed query engine over your data, no warehouse required.Get started
Nightshift

Connectors

Connect a source once. Query it everywhere.

Point Nightshift at a bucket, a warehouse, a database, or a SaaS API, and it introspects the data into a governed catalog of tables and views. People query it in notebooks, agents query it over MCP, and you never stand up a warehouse of your own.

Get started freeTalk to sales
Catalog4 sources connected
sapSDK5.7kConnected
snowflakeGuided14.2kConnected
kafkaStream31.9kStreaming
stripeSDK920Syncing
Catalog ready · 9.8M rows · the one catalog everything reads

One catalog of record

Every source becomes a governed table.

Register a connection and Nightshift introspects the schema into tables and views in one catalog. There are no pipelines to build and nothing to keep in sync by hand: the catalog is what people and agents read, and it stays current with the source.

  • Schema introspected into typed tables and views
  • One place to see what is exposed, to whom, under which policy
  • Source credentials are configured per connection, never handed to a reader
CatalogConnected
S3s3://acme-lake/ordersParquet · 2.14M rows
introspected into a governed table
orderstable
idbigint
customer_idbigint
regionvarchar
amountdecimal(12,2)
statusvarchar
created_attimestamp
A governed table over your source. Query it where it lives, or ingest the hot paths.

Where the data lives

Query it in place, or ingest the hot paths.

For object storage, Nightshift queries your Parquet and CSV where they already sit, pushing the filter and the aggregate down to the source so there is no second copy and no metered per-query compute. Ingest the tables you want faster, on the managed warehouse, whenever you choose.

  • Read object storage in place, no copy to maintain
  • The filter and the aggregate run at the source
  • Ingest hot tables onto the managed warehouse when you want speed
arr_by_regionat the edge0.41 s · no cluster
at sourcescanned and aggregated9.8M rows
over the wirereturned to the agent312 rows
The filter and the sum ran where the data lives. Only the result crossed the wire, with no warehouse in between.

Built in

Everything a source needs to be query-ready.

Any source

Object storage and cloud warehouses, relational databases, SaaS APIs like Stripe and SAP, or your own services over HTTP with auth passed through.

Snapshots and time travel

Serve a point-in-time snapshot for reproducible analysis, or the live table for work that needs the current state. Both read through the same catalog.

Governed from the first read

A connected table is access-controlled by identity from the moment it lands, so nothing is exposed until you grant it, and every read is shaped to who is asking.

The rest of the product

Nightshift Console

Connect sources, query in notebooks or over MCP, and govern every read from one place.

Learn more

Nightshift Fiber

The typed React API for building governed data apps on your catalog.

Learn more

Policy

Allow, deny, and mask down to the row, scoped to identity.

Learn more

Query

Query your governed catalog from a notebook, over MCP, or in an app.

Learn more

Connect your first source in five minutes.

Start free, register a source, and watch the catalog appear. Grant access and point an agent at it whenever you are ready.

Get started freeTalk to sales

Want to look first? Take the product tour