How AI Agents Run AutoML in h2oGPTe and Driverless AI
- Agentic AI
- Enterprise h2oGPTe
- H2O Driverless AI
This lesson explains how h2oGPTe agents combine generative and predictive AI into a single workflow. Rather than treating the two paradigms separately, an agent takes a prompt and dataset, executes an H2O Driverless AI AutoML experiment on the agent's behalf, then analyzes the completed experiment's results. The lesson walks through an example where the agent collects the artifacts Driverless AI produces, including logs and charts, and presents the finished experiment back to the user, illustrating a workflow that goes from a prompt through model building to an interpreted result.
Key takeaways
- How combining generative and predictive AI lets an agent go from a prompt to a finished analysis in one workflow
- How an h2oGPTe agent executes an H2O Driverless AI AutoML experiment based on a natural language prompt and a dataset
- What artifacts an agent collects from a completed Driverless AI experiment, including logs and charts
- Why generative AI is described as reasoning over user intent while predictive AI forecasts outcomes from data
Start owning your AI