Pipelines bring in everything your planning cycle needs: sales, stock, orders, suppliers, customers, events, POS data, ERP data, e-commerce data, or external feeds.

They clean, align, and normalize the data into a shape that dashboards and forecasts can rely on.

Pipelines don’t stop at cleaning data or refreshing dashboards.

They can send alerts, generate tasks, create purchase orders, update stock levels, or even push contacts into your CRM or marketing system.

• Generate loyalty program entries from customer behavior

• Trigger marketing campaigns when customers lapse

• Add CRM entries when high-value customers reappear

• Push signups, leads, and events directly into your workflow tools

Pipelines support any source you operate with:

• POS / ERP APIs

• Shopify / e-commerce APIs

• S3 & GCS bucket files (CSV, Parquet)

• BigQuery and ClickHouse

• Custom REST endpoints

• Uploaded files or spreadsheets

And because pipelines use a common SQL engine, you can combine sources naturally — no warehouse modeling needed.

Unlike Talend-style ETL, you don’t need a modeling layer, staging DB, or transformation DAG. Pipelines run practical, lightweight logic written for everyday operators:

• filter rows

• merge datasets

• calculate KPIs

• normalize fields

• detect anomalies

• compute planning metrics

This is where AUTOMAT’s engine becomes more than ETL. Pipelines can push data back into your operational systems:

• ERP purchase orders

• Stock adjustments

• Distribution orders (DC → stores)

• CRM / marketing entries

• Customer loyalty actions

• Webhooks to custom apps

Automation isn’t all-or-nothing.

Every step can require human review:

• Approve suggested quantities

• Override exceptions

• Confirm transfers

• Manually trigger special jobs

Teams stay in control; automation removes the routine work, not decision-making.