Puddle data warehouse

The governed data warehouse for exploration teams

Import drilling, assay, tenement, and spatial data into a PostgreSQL/PostGIS workspace with resource-scoped permissions, map-driven filters, audit, and controlled publishing into Canetoad.

Request accessView architecture
Warehouse

Data warehouse

Register schemas, tables, columns, map views, imports, filters, and layouts in one metadata-backed workspace.

PostgreSQL + PostGIS
Control

Fine-grained control

RBAC/FGA handles resources while PostgreSQL RLS remains the final tenant-data boundary.

table / filter / report
Visualize

Shared filter context

Move between server-backed grids, maps, saved filters, exports, and reports without rebuilding context.

one filter object
Publish

Publish to Canetoad

Queue scoped map layers, reports, API surfaces, or artifacts with approval, expiry, and audit trails.

publish gated
Warehouse

A warehouse built around mining data workflows

Puddle is not just file upload. It is a governed system for staging, validating, registering, and querying durable geological and spatial data.

Registry: modules, tables, columns, filters, layouts, map views.
Import buffer: preflight, chunk, stage, validate, review, commit.
Jobs: imports, exports, schemas, connector runs, publishing.
Warehouse catalogue42 registered tables
Module
Tables
Rows
Spatial
State
Drilling
4
88,057
Yes
Ready
Samples
3
1,842,930
No
Ready
Tenements
2
312
Yes
Read-only
Geophysics
5
14,880
Yes
Syncing
Preflighted
Uploaded
Staged
37 rejected
Control

Permissioning that is visible before data moves

Every resource shows what is allowed, what requires approval, and which boundary enforces access.

Resource
Analyst
DataAdmin
Boundary
drilling.collars
query
manage
RLS final
North block filter
read
share
FGA + RLS
May drill pack
view
approve
report scope
Acme SQL staging
denied
manage
connector
Audit and support trailblocking audit
09:42Priya changed table permission for drilling.collars
09:51Morgan approved schema change recovery_pct
10:08Publish job created for Canetoad map layer
10:15Support grant opened for workspace diagnostics
10:21Import buffer validation wrote 37 rejected rows
10:33Export scope preview refreshed for North block filter
10:41Role assignment synced from WorkOS
10:47API key scope rotated for publishing service
10:52Connector credentials touched by DataAdmin
Visualization

Grid, map, and filters share the same context

The map authors filters; the grid, exports, reports, APIs, and publish jobs consume the same scoped object.

Saved filters
North block, depth > 50m
attribute + geometry
Pending assay batches
attribute
depth_m >= 50
hole_type in RC,DD
geometry: polygon
          rows: 1,284
Rural satellite map with selected exploration drill points
Author on the mapDraw polygons, select in view, or paste IDs.
Consume everywhereGrid, export, report, API, publish job.
Inspect the queryCount, SRID, permission, audit outcome.
Publishing

Publish governed data into Canetoad

Publishing is a controlled destination flow: review scope, queue a job, approve the destination, and preserve the audit trail.

Registry: modules, tables, columns, filters, layouts, map views.
Import buffer: preflight, chunk, stage, validate, review, commit.
Jobs: imports, exports, schemas, connector runs, publishing.
Satellite map with exploration sampling points
Publish flow
Saved filter
Review scope
Queue publish job
Approval
Canetoad destination
Audit record
Publish job detailDataAdmin required
Job
pub_01hx9c2h
Source
drilling.collars
Filter
North block, depth > 50m
Rows
1,284
Destination
Canetoad map layer
Visibility
Canetoad internal + Acme org
Audit
aud_01hx9c5v
Built on explicit boundaries
table: drilling.collars
filter: North block, depth > 50m
fga: table:query allowed
rls: org_id = acme, workspace_id = pilbara
publish: approval_required
audit: blocking
PostgreSQL/PostGIS
Authoritative data plane.
Custom Data Registry
Registry for tables and filters.
Fine-Grained Authorization
Resource-scoped authorization.
Job Pipelines
Imports, exports, publish artifacts.

Bring your exploration data under control

Start with a governed workspace, import real project data, and publish only scoped outputs your team has approved.

Request accessRead architecture