Pipeline Features

Overview

This page covers every feature available in the Pipeline module. Use it as a reference when building, monitoring, or troubleshooting your pipelines.

Pipeline Statuses

A pipeline can be in one of three states:

Status	Description
Draft	Pipeline has been created but not yet activated. Workflow is editable and test runs are available.
Active	Pipeline is running on its defined schedule. Outcomes are continuously refreshed.
Paused	Pipeline is stopped. The last outcome is preserved but not updated. Resume when ready.

The Overview Tab

When you open a pipeline, the Overview tab gives you a health summary at a glance.

It shows:

Runs — total number of executions
Success — percentage of successful runs (e.g., 100.0%)
Failed — number of failed runs
Pipeline Prompt — the natural language description used to generate the workflow (read-only)
Recent Runs — the last 5 runs with timestamps and status
Data Tables — the named Pipeline Outcomes produced by this pipeline

The Workflow Tab

The Workflow tab shows the pipeline’s logic as a visual directed graph.

Pipeline Prompt

The top section shows the natural language description used to generate (and optionally regenerate) the workflow. In Draft state this is editable; in Active/Paused state it is read-only.

Workflow Graph

The graph displays every step of the pipeline as a connected sequence of nodes. Nodes are color-coded by type:

Node Type	Color	Purpose
TOOL	Purple	Calls an external tool to fetch or act on data
READ DB	Blue	Reads records from a connected database or data store
TRANSFORM	Teal	Reshapes, renames, or processes data fields
FILTER	Pink	Applies conditional logic to include/exclude records
GROUP	Pink/Accent	Groups and aggregates records (sum, count, etc.)
WRITE	Green	Writes the final output to the Internal Data Store

Clicking a node opens a detail panel showing its configuration in Fields view or raw JSON view.

Controls

Dev Mode toggle — switch to a raw view of the workflow definition (for advanced inspection)
Test Run — execute the pipeline immediately against real data (in Draft or Active state)
Activate / Pause / Resume — change the pipeline’s running state

Use the Test Run feature after any change to the pipeline description to validate output before it runs on schedule.

The Runs Tab

The Runs tab provides a full history of all pipeline executions.

Each row in the Run History table shows:

Run ID — unique identifier for the execution
Started — date and time the run began
Duration — how long the run took (typically 1–3 seconds for well-formed pipelines)
Status — success or failed

Runs are paginated (10 per page). Use the page controls at the bottom to navigate through history.

The pipeline overview stat card shows total runs, success rate, and failure count. If your success rate drops below 100%, investigate recent runs for errors and check that your connector credentials are still valid.

The Data Tab

The Data tab shows the current Pipeline Outcomes — the structured output tables produced by the most recent successful run. Pipeline Data tab showing deals_over_10k output table labeled Pipeline output

Pipeline Data tab showing deals_over_10k output table labeled Pipeline output

Each outcome table is labeled with its name (e.g., deals_over_10k, aggregated_expenses) and shows records as rows. If the pipeline has never run or is paused, the message “No records yet — Records will appear here after the pipeline processes data” is shown.

The Settings Tab

The Settings tab gives you full control over the pipeline configuration.

Configuration Fields

Field	Description
Name	The pipeline’s display name
Description	The natural language description / pipeline prompt
Source	The Connector providing input data
Destinations	Where output data is stored (always Internal Data Store)
Schedule Type	Manual or Cron
Cron Expression	The cron schedule (e.g., `0 9 * * *` = daily at 9:00 AM UTC)
Schedule	Human-readable schedule (e.g., “At 09:00 AM (UTC)“)
Max Consecutive Failures	Number of failed runs before the pipeline auto-pauses (default: 3)
Created	Timestamp and user who created the pipeline
Last Updated	Timestamp and user of the most recent change

Auto-Pause on Failure

The Max Consecutive Failures setting (default: 3) automatically pauses the pipeline if it fails repeatedly. This prevents a misconfigured pipeline from silently running and producing bad data.

If your pipeline auto-pauses due to consecutive failures, check the Runs tab for error details, fix the root cause (usually a connector issue or schema change), then click Resume.

Scheduling Options

Pipelines support five scheduling modes:

Option	Cron equivalent	Typical use
Manual	(none)	On-demand runs, testing, one-off jobs
Hourly	`0 * * * *`	Live CRM data, monitoring dashboards
Daily	`0 9 * * *`	Financial reports, daily summaries
Weekly	`0 9 * * 1`	Weekly rollups, lower-frequency aggregations
Custom	Any cron expression	Specific timing requirements (e.g., every 15 min, weekdays only)

The Pipelines Overview Dashboard

At the top level of the Pipelines section, the Overview tab shows aggregate health metrics across all pipelines in your project.

Metrics shown:

Active Pipelines — how many are currently running (vs. total)
Total Runs — all executions across all pipelines
Success Rate — aggregate success percentage
Avg Runs / Pipeline — average executions per pipeline
Run Health by Pipeline — bar chart per pipeline showing success vs. failure ratio
By Source — breakdown of runs by connector type

Next Steps

Using Pipelines in Experiences — display Pipeline data in HITL dashboards
Best Practices — tips for building reliable, maintainable pipelines

Home

Workflows & Use Cases

Features

Authentication

Overview

Pipeline Statuses

The Overview Tab