What's Inside a Driverless AI AutoDoc Report
- Observability
- H2O Driverless AI
This lesson shows how to download the AutoDoc report that Driverless AI automatically generates for a completed experiment, and walks through what is inside it. The document opens with the experiment name, user, and date, followed by a table of contents covering the dataset overview, model tuning, feature evolution and transformation, the best predictive model, and additional plots. It also covers descriptive statistics for numeric and categorical features and a deployment chapter explaining the Python and MOJO scoring pipeline options, giving teams one document for explainability and regulatory visibility.
Key takeaways
- How to download the AutoDoc report from a completed experiment's status page
- What sections an AutoDoc report includes, from the dataset overview to the final model pipeline
- Why AutoDoc matters for companies that need explainability and visibility into how a model was built
- How the deployment chapter of AutoDoc explains the Python and MOJO scoring pipeline options
Start owning your AI