Transformations (Edge)

Transformations is a built-in Hevo tool that converts raw data in your Destination into clean, structured tables for reporting and analysis. Data loaded by Pipelines typically arrives as unprocessed tables that are rarely in a format suitable for dashboards, reports, or analysis. Transformations lets you clean, reshape, and structure that data directly in your Destination.

Built on top of dbt, Transformations gives your data team a single place to write dbt models, run them against your Destinations, and track every execution with detailed logs.

With Transformations, you can:

• Connect your GitHub account so Hevo can access the repository where your dbt models are stored.

• Run dbt models against your Destinations to clean and reshape raw data into structured tables.

• Schedule runs to happen automatically, trigger them after a Pipeline sync, or run them on demand.

• Track every execution with step-by-step logs to monitor progress and troubleshoot issues.

---

Transformations in the ELT Workflow

The following diagram illustrates how Transformations fits into the ELT (Extract, Load, and Transform) workflow:

In the ELT process, the Extract and Load steps are handled by Pipelines, which pull data from the configured Sources and load it into your Destination. Transformations handles the final Transform step by running dbt models to convert that raw data into clean, structured tables for your reporting and analytics tools.

---

Core Concepts

The core structure of Transformations defines how dbt projects are organized and executed within Hevo. It consists of the following:

• Transformation Project: The top-level container that links a GitHub repository to Hevo. It contains all the environments, jobs, and run history for the connected repository.

• Environment: The configuration that defines where your dbt models run. Each environment points to a specific Destination and Git branch. It can be of type staging (for testing changes) or production (for live data).

• Job: A configured task that defines which dbt commands to run, in which environment, and on what schedule or trigger.

• Run: A single execution of a job. Each run records the outcome, step-by-step execution logs, lineage, and dbt artifacts.

---

Key Features

Transformations provides the following features:

GitHub integration

Your dbt models stay in your GitHub repository. Each time a job runs, Hevo pulls the latest version of your code automatically, so runs always execute against the current version of your models.

Staging and Production environments

Each Transformation project supports these two environment types. Staging provides a safe space to test and validate code changes before they go live, while production runs dbt models on your actual business data.

Environment variables

Values stored outside your dbt code that control how your models behave at runtime. You can set different values for each environment, so the same dbt code runs differently in staging and production environments.

Post Pipeline sync trigger

A deploy job can be configured to trigger automatically when selected Pipelines finish syncing, so your dbt models always execute on the latest data without any manual steps.

Monitoring and lineage

Every job run produces a step-by-step execution log that shows exactly what happened at each stage. A lineage graph maps the relationship between your Source data, dbt models, and final output tables. This makes it easy to trace where data comes from and how it was processed. Logs are searchable and can also be downloaded for further review.

Single platform

Transformations is a Hevo tool, which means your data ingestion and dbt model execution are managed from the same platform. You do not need to configure or maintain a separate tool.

---

Supported Destinations

Transformations currently supports the following Destinations:

• Amazon Redshift

• Google BigQuery

• Snowflake

---

When to Use Transformations

You can use Transformations to:

• Convert raw data loaded by Pipelines into clean, structured tables ready for reporting and analysis.

• Automate dbt model execution with scheduled runs, post-Pipeline triggers, or on-demand runs.

• Test code changes safely in a staging environment before they affect your production data.

• Maintain a consistent, auditable record of all changes to your dbt models across your team.

---

When Not to Use Transformations

Transformations may not be the right fit in the following scenarios:

• Your dbt project is hosted on GitLab, Bitbucket, or any provider other than GitHub. Transformations currently supports GitHub repositories only.

• You need your dbt models to run before data is loaded into your Destination. Transformations runs the models only after data has been loaded.

• You need real-time data transformation. Transformations runs on a schedule, on demand, or after a Pipeline sync, not as data arrives.

• You are using a Destination other than Amazon Redshift, Google BigQuery, or Snowflake. Transformations currently supports only these three Destinations.

---

See Also

• Understanding Transformations

• Working with Environments

• Working with Jobs

• Managing Transformation Projects

• Managing GitHub Accounts

• Monitoring and Managing Runs

---

Revision History