Seek Build Tools (SBT) Overview
Seek Build Tools (SBT) is a framework that empowers data scientists and analysts to design, develop, and deploy complete data science workflows, from data ingestion to interactive visualization, within a single project. Once completed, the project can be deployed to Insight Cloud, Seek's cloud-based analytics platform, where users can securely run your app with their own data.
ℹ️ Three terms that sound alike but mean different things. Getting these straight prevents most SBT confusion:
The SBT package is the command-line tool you install (current versions are numbered 1.0.x, like 1.0.10). Everyone uses the same tool, regardless of which kind of app they're building. "Upgrade your SBT" means reinstalling this package with a newer pinned version.
An SBT Beta app (AppModel) is an app built on the legacy framework. The D&A apps running in production today are Beta apps, built and managed with the
--betaflag.An SBT 1.0 app (AppSpec) is an app built on the new framework, which runs in containers instead of Snowflake stored procedures.
So "the SBT 1.0 migration" means converting apps from Beta to 1.0, not upgrading your command-line tool. Both frameworks run side by side during the migration; Beta apps keep working unchanged, and new apps should target SBT 1.0. Across the Creator guides, SBT 1.0 changes are called out explicitly; anything without a version note works the same in both frameworks. Version notes like "SBT 1.0.7+" refer to the package (term 1).
Why SBT?
SBT simplifies the process of building and deploying complex data analytics apps by providing a consistent interface for creating modular data pipelines while abstracting away the underlying infrastructure. This allows you to focus on developing your app's logic and visualizations, rather than managing the tools and services needed for production.
In addition to a streamlined development experience, SBT includes features that support common data engineering and data science workflows, such as schema validation, data versioning, data lineage tracking, machine learning model management, and more.
SBT also enables you to publish your apps to Insight Cloud. From there, an app can be installed directly into the organizations you choose. Or, if your organization is set up as a publisher, offered on the Insight Cloud App Store to reach a broad audience of data analytics professionals.
What is SBT?
SBT is a command-line tool (CLI; you use it by typing commands in a terminal) and Python toolkit designed for efficiently building data pipelines and deploying them to the Insight Cloud. It provides a set of commands, classes and conventions that help you structure your project, define data sources, write models, create visualizations, and package everything into a deployable app.
The workflow at a glance
Building an app follows the same path every time:
Set up: install SBT and configure your project. See Setting Up SBT.
Build: define sources, write models, and run them locally to validate. The Quickstart walks you through your first app step by step.
Visualize: build a Sigma dashboard on your model outputs. See Data Visualizations with Sigma.
Publish: bundle your models and dashboard into an app and publish it to Insight Cloud, then install it directly into organizations or offer it through the App Store as a publisher. See App Publishing Lifecycle.
AppModel vs AppSpec at a glance
Converting an app from Beta to 1.0 is covered in the SBT 1.0 App Migration Guide.