Export to Data Warehouse

Git AI Enterprise supports exporting raw telemetry to a data warehouse, making it possible to join AI coding data with existing engineering, finance, and security datasets. Build custom dashboards, run ad-hoc queries, and feed AI adoption metrics into existing reporting pipelines.

Data warehouse export is available on Enterprise plans (Cloud and Self-Hosted).

Supported Warehouses

Exported Data

Exports include all telemetry collected by Git AI, matching the tables available in the Data Catalog. Key datasets include:

• Agent sessions — session metadata including token usage, cost, duration, and model
• Pull request metrics — AI code percentages, churn rates, and review outcomes per PR
• Contributor metrics — per-developer AI usage patterns and adoption trends

Exports run on a configurable schedule. Each export contains only new and updated records since the previous run.

Configuration

Configure warehouse exports from Settings > Export in the Git AI dashboard. Each warehouse requires connection credentials and a destination location.

Snowflake

Snowflake exports use key-pair authentication. Generate an RSA key pair and assign the public key to a Snowflake user with write access to the target schema.

To configure Snowflake:

1. Create a dedicated database and schema (e.g. git_ai.telemetry)
2. Create a Snowflake user for the export
3. Assign the RSA public key to the user
4. Grant USAGE on the compute warehouse and ALL on the target schema to the user's role

Databricks

Databricks exports write to a Unity Catalog volume as Delta tables. Provide a personal access token with write access to the target catalog and schema.

To configure Databricks:

1. Generate a personal access token with workspace and catalog write permissions
2. Create or select a Unity Catalog and schema for Git AI data
3. Enter the workspace URL, token, catalog, and schema in Settings > Export

BigQuery

BigQuery exports use a Google Cloud service account. Create a service account with the BigQuery Data Editor role on the target dataset, then upload the JSON key file.

To configure BigQuery:

1. Create a Google Cloud service account (e.g. git-ai-export)
2. Grant the BigQuery Data Editor role on the target dataset
3. Generate a JSON key for the service account
4. Upload the JSON key file in Settings > Export

Export Schedule

Exports run on a configurable schedule. Set the interval in Settings > Export after configuring a warehouse connection. The default schedule is every 6 hours. Available intervals:

• Every hour
• Every 6 hours
• Every 12 hours
• Daily

Each export is incremental — only records created or updated since the last successful export are included.

Monitoring Exports

The Settings > Export page displays the status of each configured export:

Failed exports include an error message with details. Common issues include expired credentials, insufficient permissions, or network connectivity problems. Resolve the issue and trigger a manual re-run from the dashboard.

Self-Hosted Exports

Self-hosted deployments support the same warehouse destinations. The Git AI instance must have network access to the target warehouse. For Snowflake and Databricks, ensure outbound HTTPS access from the Git AI host. For BigQuery, the instance needs access to the BigQuery API (bigquery.googleapis.com).

See Self-Hosting Git AI for network configuration details.