Snowflake + Go Fig
Pull data directly from your Snowflake data warehouse into Go Fig for unified financial analytics and reporting.
Snowflake is where most mid-market data teams already centralize operational data. Go Fig connects natively, querying tables in place or pulling them into the Financial Intelligence Graph alongside your accounting ledger and CRM. Celeste can reach across Snowflake objects without anyone exporting CSVs or rebuilding logic in a BI tool.
Key facts
- Auth
- Key-pair (JWT), OAuth, or password
- Access mode
- Query-in-place or replicated
- Zero-copy
- SHARE-based access supported
- Warehouse control
- Dedicated warehouse selectable
- Sync cadence
- Live query or scheduled materialization
What you can do with Snowflake data in Go Fig
Unified Financial Reporting
Combine Snowflake operational data with ERP, CRM, and payments for blended margin, CAC, and segment views.
Live Query Without Movement
Query Snowflake in place where data residency or governance prevents replication, with results joined to other Go Fig sources.
Historical and Time-Travel Analysis
Use Snowflake's Time Travel for retroactive close adjustments and trend analysis on prior states of the same table.
Data available from Snowflake
Go Fig extracts and normalizes the following data from your Snowflake account:
How to connect Snowflake
Create a Go Fig user and role in Snowflake
Provision a dedicated user with key-pair authentication enabled. Create a role granting USAGE on the warehouse, database, and schema, plus SELECT on the tables and views you want exposed. Avoid granting ACCOUNTADMIN or SYSADMIN; least-privilege roles work for read-only sync.
Choose access pattern: replicate or query-in-place
Replication pulls data into Go Fig's managed storage and is best for joining with other sources at low latency. Query-in-place leaves data in Snowflake and pushes filtered queries down on demand, which is the right choice for very large tables, governed datasets, or when egress is restricted. SHARE-based zero-copy access is also supported.
Pick a warehouse and concurrency settings
Point Go Fig at a dedicated XS or S warehouse so its compute cost is auditable and does not contend with your data team's analytics workloads. Auto-suspend after 60 seconds keeps idle cost near zero. The connector dashboard reports per-query credit usage so finance can see exactly what reporting is costing.
Schedule sync or query cadence
Replicated tables can be refreshed on schedules from 5 minutes up to daily. Query-in-place views are evaluated at request time, with optional caching layers in Go Fig to avoid repeat warehouse spin-up. Time Travel is used for backfills when you need to reproduce a prior state of a table for an audit-friendly close.
Authentication: Key-pair authentication (JWT) is the recommended production path. OAuth 2.0 via an external IdP is supported for SSO-controlled environments. A dedicated read-only role on a chosen warehouse keeps Go Fig's compute cost visible and isolated from your other workloads.
Common Questions About Snowflake Integration
Does Go Fig replicate Snowflake data or query in place?
Both are supported. Most teams replicate the tables they will join repeatedly with accounting and CRM data, because that delivers the lowest-latency joins. Very large tables, sensitive datasets governed by data-residency rules, or tables that change too rapidly for batch refresh are typically left in Snowflake and queried in place. You can mix the two patterns inside a single connection.
Which authentication methods are supported, and which is recommended?
Key-pair authentication (JWT) is recommended for production because it avoids long-lived passwords and integrates cleanly with Snowflake's network policies. OAuth 2.0 via an external IdP is supported for organizations standardized on SSO. Password auth is supported but discouraged because Snowflake itself is steering customers off it. MFA is honored at whichever auth path you choose.
How does Go Fig manage Snowflake compute cost?
Sync runs against a warehouse you specify, so cost is fully attributable. We recommend a dedicated XS or S warehouse with 60-second auto-suspend, which keeps idle cost near zero and isolates Go Fig's queries from your data team's analytics workloads. The connector dashboard shows per-query credit usage so finance has a clear view of what unified reporting is actually costing in Snowflake credits.
Can Go Fig use Snowflake Secure Data Sharing (SHARE)?
Yes. If you prefer to expose data via a SHARE rather than granting Go Fig direct table access, the connector mounts the share as a read-only inbound database. This is a common pattern for regulated organizations because no data leaves your Snowflake account, and the share can be revoked instantly if you change vendors. Zero-copy means there is no storage duplication on your side.
How are schema changes and column additions handled?
Go Fig polls INFORMATION_SCHEMA on a schedule and detects new columns, dropped columns, and type changes. New columns appear automatically in Go Fig with the next sync. Type changes that would break downstream flows raise a notification rather than silently casting, so your reporting does not quietly start producing wrong numbers when an upstream team alters a table definition.
Industries running Snowflake with Go Fig
Strategic CFOs in these industries typically stitch Snowflake into their Financial Intelligence Graph alongside their ERP and operational systems.
Ready to connect Snowflake?
See how your Snowflake data looks in Go Fig with a personalized demo.
Book a Demo