Skip to main content

What is AI App Store?

The AI App Store is the unified interface for H2O AI Cloud, giving you a single place to find and run AI apps, train and deploy models, build ML pipelines, manage features and data, and handle day-to-day operations. You can also build your own apps with the H2O Wave SDK, and the platform also includes built-in tools for MLOps, Feature Store, Workflows, AI Engines, and Notebooks.

Problems it solves​

Most data science projects end with software that helps people make decisions, or makes those decisions automatically.

These tools help people choose better by showing them the right information in a format they can understand. The team that builds the tool already handles the hard parts: which data to use, which algorithms to apply, and how to present the results.

Building and shipping these tools is challenging due to a few reasons:

  • Infrastructure: Machine learning needs significant storage and compute, and adding it to a software project is much more complex than building conventional software.
  • Talent: You need a mix of specialists working together: data scientists, data engineers, backend and frontend engineers, and IT/ops, all in close contact with the people who will use the result.
  • Time to market: Requirements change constantly as markets, competitors, and customer expectations shift. Teams no longer have months or years to ship. They need to prototype fast, get early feedback, and iterate or move on.

In short, a team has to invest significant effort wiring up data, libraries, tools, and infrastructure before they can focus on what matters most: delivering results to users. H2O AI Cloud handles that wiring for you, and the AI App Store is your entry point.

Capabilities​

The AI App Store covers the full ML lifecycle behind one UI and one API. From the left navigation you can:

  • Discover and run apps: Browse the catalog, start your own private app instance, and share it with your team. See Using apps and Concepts.
  • Train and deploy models: Track, version, and serve models from the built-in MLOps UI, and launch Driverless AI, H2O-3, or LLM Studio through AI Engines.
  • Build and run ML pipelines: Write and schedule Workflows, and work in hosted Notebooks without leaving the platform.
  • Manage features and data: Define and serve features in the Feature Store, and share files through Drive.
  • Collaborate in Workspaces: Keep each team's apps, models, secrets, files, and permissions separate.
  • Build custom applications: The Wave SDK lets data scientists and data engineers build real-time analytical web apps in pure Python, with no JavaScript, HTML, or CSS required. See the App developer guide.
  • Operate at scale: Manage authorization, audit activity, monitor usage, and configure the environment. See the Admin guide.

For more on how these pieces fit together, see the H2O AI Cloud architecture documentation.

Feedback