Jupyter Notebooks

This page includes demo Jupyter Notebooks for researchers and practitioners who want to try out or better understand the Digital Twin Simulation framework.

Full Pipeline Demo

This notebook walks through the full pipeline of running a Digital Twin Simulation. It’s the most comprehensive example and covers:

  • Downloading the dataset

  • Processing personas and questions

  • Running LLM simulations

  • Converting results to CSV format

  • Running accuracy evaluations

  • Performing pricing analysis

Link: demo_full_pipeline.ipynb on GitHub

Prerequisites:

  • Remember to update the PROJECT_ROOT_PATH in the notebook to match your local repository.

  • You must first clone the repository to your local workspace.

Simple Simulation Demo

This notebook shows how to run a basic simulation using a small set of custom questions. It’s a quick and easy guide for adding your own questions to the simulation pipeline with minimal setup.

Link: demo_simple_simulation.ipynb on GitHub

Prerequisites:

  • Remember to set up your own LLM API key.