OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Research Agents

Curie

Curie is a Python research experimentation agent that automates computational scientific workflows, including experiment planning, implementation, execution and result analysis.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Curie — OpenViST project profile in Research Agents.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Curie when a research question can be investigated through computational experiments and you can provide the question, model credentials, and any relevant data or code.

Who it is for

  • Researchers who want to explore a scientific question through computational experiments.
  • Machine-learning practitioners who want to test strategies on a dataset, including with their own starter code.

Documented uses

  • Turn a research question into an experiment workflow and review its results.
  • Compare machine-learning methods, such as robustness to added data noise.
  • Explore a dataset using optional starter code and generate experiment reports and process artifacts.
Before you choose

Requirements & limits

  • Docker is a stated installation prerequisite, alongside a Python package or manual developer installation path.
  • The quick-start example requires model configuration and an API key; it shows a Claude model with an Anthropic key.
  • Experiments may take several minutes depending on task and budget; the README recommends a terminal multiplexer and advises against Jupyter Notebook.
  • A GPU is recommended for machine-learning training tasks.
  • The repository identifies its license as Apache-2.0.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install Curie and prepare an experiment

Set up the documented Python package and prepare the inputs and model credentials used in Curie's quick-start example.

Before you start

  • Docker installed
  • Python environment for the package
  • A model identifier and matching API key; the README example uses a Claude model and an Anthropic API key

Steps

  1. Install Docker and follow the README's Docker socket permission and docker ps verification instructions.

  2. Install the package with pip install curie-ai, or use the repository's manual developer installation instructions.

  3. Set up the API key configuration by referring to curie/setup/env.sh.example.

  4. Prepare a research question; for dataset work, provide a dataset directory and optionally a codebase directory and requirements file.

Expected outcome

Curie's documented prerequisites and inputs are prepared for calling its experiment function.

Guide limitations

  • This guide covers setup and input preparation, not a complete experiment run.
  • The README advises against using Jupyter Notebook and recommends tmux or a similar terminal multiplexer for experiments that may take several minutes.
  • A GPU is recommended for machine-learning training tasks.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
13 October 2026
Original website title
Just-Curieous/Curie · Official homepage title · retained in its source language
How we review sourcesSuggest a correction