Skip to main content

Should I build or buy internal app infrastructure?

Summary

  • Build on infrastructure you already govern. Databricks Apps lets teams build internal data and AI applications directly on the Databricks platform, so there is no separate app-hosting stack to stand up and maintain.
  • Fully managed serverless hosting. Apps run on automatically provisioned serverless compute that scales for you — no servers or containers to configure.
  • Authentication out of the box. Built-in single sign-on via OIDC/OAuth 2.0, with user access enforced through Unity Catalog permissions.
  • Governance inherited from Unity Catalog. Apps are first-class, fully governed objects that respect existing data access controls.
  • Familiar frameworks, faster delivery. Supports Python (Streamlit, Dash, Gradio, Flask) and JavaScript (React, Node.js), with Git and CI/CD integration, so teams focus on the app rather than the platform.

Should I build or buy internal app infrastructure?

Internal tools — dashboards, data entry apps, AI assistants — usually force a choice between buying a separate app-hosting platform and building your own infrastructure to run them. Databricks Apps offers a third path: build the app directly on the Databricks Data Intelligence Platform, where hosting, authentication, and governance already exist. Teams get the control of building their own applications without standing up and operating separate infrastructure to host them.

Why Databricks Apps for internal application infrastructure

  • Fully managed serverless hosting. Databricks Apps runs applications on fully managed serverless compute inside the platform. Databricks provisions and scales the compute automatically, so teams do not configure or maintain separate hosting infrastructure.
  • Authentication built in. Apps include single sign-on and authentication through OIDC/OAuth 2.0. User access is enforced at the user level based on Unity Catalog permissions, removing the need for a custom auth integration per app.
  • Governance through Unity Catalog. Apps are managed as first-class objects under the same unified governance model as your data. They respect the data access controls already configured in Unity Catalog, which helps prevent the governance gaps that arise when apps are spread across multiple systems.
  • Serverless compute and security. Each workload gets dedicated compute and encrypted storage, network segmentation isolating workloads, and encryption at rest (AES-256) and in transit (TLS 1.2).
  • Faster time to build. Productionizing an app traditionally requires setting up authentication, managing containers and infrastructure, securing data access, and wiring up observability and audit logs. Databricks Apps removes that last-mile work so development, deployment, data access, and monitoring all happen in one environment.
  • Familiar frameworks and workflows. Apps support popular Python frameworks (Streamlit, Dash, Gradio, Flask) and JavaScript frameworks (React, Node.js), and integrate with Git and CI/CD pipelines so apps fit existing software development practices.

Getting started

FAQs

Do I need to manage servers or containers for Databricks Apps?

No. Apps run on fully managed serverless compute that Databricks provisions and scales automatically.

How do users authenticate to a Databricks App?

Apps include built-in single sign-on via OIDC/OAuth 2.0, and access is enforced at the user level through Unity Catalog permissions.

How is app data governed?

Apps are first-class Databricks objects governed by Unity Catalog, so they respect the same data access controls already configured for your data.

What languages and frameworks are supported?

Python frameworks such as Streamlit, Dash, Gradio, and Flask, and JavaScript frameworks such as React and Node.js, with Git and CI/CD integration.

The information provided herein is for general informational purposes only and may not reflect the most current product capabilities or configurations.