numerai-research

An end-to-end Numerai research workflow orchestrates experiment design, model implementation, automated training runs, and report generation into a single automated cycle. It sequences multiple skills to deliver runnable experiments and optional deployment artifacts.

Yes, automated research workflows can produce comprehensive experiment.md reports containing standard plots and tracking metrics. This allows you to compare multiple experiment configurations visually and evaluate model performance systematically.

No, a complete research workflow can optionally package a deployable Numerai pickle for you. It handles the full deliverable packaging automatically, including uploading the pickle via MCP if requested.

The best way to run iterative Numerai experiments is to use an orchestrated workflow that designs, runs, and compares multiple configurations. It automatically tracks metrics and generates comprehensive visual reports for each run.

Yes, the workflow enforces a sequence that includes implementing new model types if required by your experiment design. This ensures custom models are integrated before running scout and scale experiments.