MongoDB + Go Fig

Database

Connect MongoDB to Go Fig for analytics on your NoSQL document data without complex ETL.

MongoDB stores your application data, but querying it for business analytics is a different challenge. Go Fig connects directly to MongoDB, letting you analyze document data alongside relational sources, spreadsheets, and APIs. No need to flatten your data or build a separate data warehouse.

Connect MongoDB See All Integrations

Key facts

Deployments: Atlas, self-hosted, VPC
Sync mode: Change streams (real-time)
Schema handling: Inferred + override
Auth: SCRAM, X.509, AWS IAM

SOC 2 Type II · All integrations

What you can do with MongoDB data in Go Fig

Application Analytics

Analyze user behavior, feature usage, and application performance from MongoDB data.

Cross-Source Reporting

Combine MongoDB application data with CRM, billing, and marketing data.

Operational Monitoring

Track key business metrics stored in MongoDB with real-time dashboards.

Data available from MongoDB

Go Fig extracts and normalizes the following data from your MongoDB account:

Collections

Documents

Nested objects

Arrays

Aggregations

Time series data

User records

Transaction logs

Application events

Custom schemas

How to connect MongoDB

Create a read-only MongoDB user

In Atlas or self-hosted MongoDB, create a database user with the read role on the databases Go Fig should access. For Atlas, also add Go Fig's egress IPs to the Network Access allowlist (or use VPC peering for production).

Connect via connection string

Provide the standard mongodb+srv:// connection string. Go Fig connects, lists databases and collections, and samples a few hundred documents per collection to suggest a flattened schema.

Tune the schema flattening

MongoDB documents nest. Go Fig auto-flattens common patterns (e.g., address.city becomes address_city) and unwinds arrays into separate rows where useful. You can override per-field to keep certain nested objects as JSON for downstream parsing.

Pick incremental sync key

For collections with high write volume, configure incremental sync on a timestamp or ObjectId field. This avoids full collection scans on every refresh and keeps load on production low.

Authentication: MongoDB connection string with read-only user credentials. Atlas customers can use Atlas's IP allowlist; self-hosted users can restrict by source IP or run Go Fig's connector inside their VPC. SCRAM-SHA-256, X.509, and AWS IAM auth are all supported.

Common Questions About MongoDB Integration

How does Go Fig handle MongoDB's flexible schema?

Documents in the same collection can have different fields. Go Fig samples documents to infer a union schema and flags fields that appear in fewer than 100% of documents. Schema drift between syncs (new fields appearing) is detected and surfaced rather than silently ignored, so downstream reports don't break.

Can I query nested objects and arrays directly?

Yes. Nested objects are flattened with dot-path naming (user.address.city → user_address_city). Arrays can be unwound (one row per array element) or kept as JSON columns depending on what makes sense for analysis. The choice is per-field, not collection-wide.

Does Go Fig support MongoDB Atlas, self-hosted, and AWS DocumentDB?

Atlas and self-hosted MongoDB are fully supported. AWS DocumentDB and Azure Cosmos DB (Mongo API) work for most use cases since they implement the MongoDB wire protocol, though edge-case features like change streams behave differently. We can confirm fit on a 15-minute call.

Will Go Fig hit my production database?

By default, yes, with a read-only user and incremental sync to minimize load. For high-traffic production clusters, the recommended pattern is to point Go Fig at a read-replica or analytics node. Atlas customers typically dedicate an M10+ analytics node to BI queries; Go Fig works the same way as any other BI tool against that node.

How does Go Fig handle ObjectIds and BSON-specific types?

ObjectIds are stored as their 24-character hex string representation, which preserves the embedded timestamp and lets you join cleanly to other systems that reference Mongo IDs. Decimal128, Date, and other BSON types are converted to standard SQL types (DECIMAL, TIMESTAMP) without precision loss.