Skip to main content

​
Overview

Flows are visual data pipelines that transform, clean, and analyze your data without writing code. You add nodes to a canvas and connect them to build reusable transformations, from a simple filter to a multi-source analysis. Think of a Flow as:
  • A visual query builder that does not require SQL knowledge
  • A reusable data transformation recipe
  • Automated data preparation for your dashboards and reports

​
How Flows Work

A Flow is a chain of nodes laid out top to bottom. Data enters from an input node, passes through processing nodes, and lands in an output node.
Each node receives data from the node above it, performs one step, and passes the result down. Click a node’s preview to see its output at that point.
A Flow canvas with a top-to-bottom chain of nodes: Get Data, Transform, and Save Table.

​
Creating a New Flow

1

Open your project

From the left sidebar, click Projects and select the project where you want the Flow.
2

Create the Flow

Click + Add Flow. The Flow opens on the canvas named “Untitled Flow”. Rename it from the top bar.
3

Add your first node

Click + Add node, then choose Get Data to pull in a table to start from.

​
The Node Palette

Click + Add node to open the palette. It is a searchable popover, so you can type a name or what you want to do (try “join” or “pivot”) to find the right node. Nodes are grouped into five sections:
The Add node palette popover with a search box and node tiles grouped into Input, Process, Merge, Output, and Tools.
Input
  • Get Data - pull rows from a connected table or Virtual Table
  • Import Data - upload a CSV or spreadsheet straight into the Flow
Process
  • Transform - filter, sort, group, aggregate, pivot, and add fields
  • AI Enrichment - use AI to add, classify, or extract fields
  • SQL/Python - run a custom SQL query or Python script
  • Forecast - project future values from a time series
  • Segment - group rows into clusters by similarity
Merge
  • XLOOKUP - join two tables on a matching key
  • Append - stack rows from multiple tables
Output
  • Save Table - persist the result as a Virtual Table
  • Chart - render a chart from the data
  • AI Summary - narrate the data with an AI summary
Tools
  • Note - add a markdown note to the canvas
  • Scenarios - define what-if scenarios with named parameters
Most data shaping happens in the Transform node. It combines what used to be separate filter, sort, group, pivot, and formula nodes into one place.

​
Connecting and Configuring Nodes

To connect two nodes, drag from the handle at the bottom of one node to the top of the next. To configure a node, click the pencil icon to open its modal, set it up, and save. To remove a connection, click the line and press Delete.
The Transform node configuration modal open, showing the field list and selected fields.

​
Previewing Data

Click the preview icon on any configured node to see its output up to that point.
A node's preview panel showing the output rows and columns at that step of the Flow.
Previews run on a sample of your source data to keep things fast. If a node returns 0 rows in the preview, a banner notes the preview used a limited sample and the full dataset may still match. Run the Flow to see results across all your data.

​
Running Flows

Click Run in the top bar to execute the whole Flow. You get a notification on the bottom left when it finishes. If you’ve connected your Slack account to Go Fig, you’ll also receive a direct message in Slack when a flow completes or errors. To run on a schedule, open Settings and set a recurring schedule, then turn on the schedule toggle.
If your Flow has email notifications configured, Run asks whether to send emails for this run. Scheduled runs always send automatically.
Go Fig auto-saves your Flow as you work.

​
Performance Tips

Filter early: Filter rows in your first Transform node so downstream steps process less data. Go Fig pushes filters back to your source when possible, which can speed up a Flow significantly.
Keep only the fields you need: Drop unused fields in the Transform Fields tab to reduce memory and speed up processing.
Parallel AI: When several AI Enrichment nodes process the same source data, Go Fig runs them in parallel automatically.
Query timeouts: Queries on external databases (Snowflake, BigQuery, Azure SQL, PostgreSQL) time out after 90 seconds. Queries on your uploaded files and synced connector data time out after 5 minutes. In both cases, filter early, select only the fields you need, and break complex transformations into smaller steps.