Build and Deploy Apps with H2O Wave
- Generative AI
This lesson walks through model consumption using H2O's open-source Wave library, which lets you build multi-page applications and chatbots in Python alone. It demos a System Prompt app that runs two chat interfaces side by side, comparing how a large language model responds with and without a defined system prompt and persona. It shows how the same Python code adapts its layout for mobile screens, then walks through packaging and deploying the app to H2O's AI Cloud using a single CLI command. It closes by showing the App Store view where admins manage running app instances and visibility settings.
Key takeaways
- How H2O Wave lets you build a multi-page chatbot application using only Python, no JavaScript or CSS required
- What a system prompt changes in a large language model's response, shown side by side in a demo app
- How the same Wave application code can define a different layout for mobile versus desktop screens
- Why a single CLI bundle command packages, deploys, and streams logs for an app running in H2O's cloud
- The App Store admin view shows running instances, visibility settings, and tags for each deployed app
Start owning your AI