Metadata-Version: 2.4
Name: nucleai
Version: 0.0.1
Summary: AI support for fusion experiments - feature extraction, embeddings, and natural language data analysis
Author-email: Simon-McIntosh <simon.mcintosh@iter.org>
Requires-Python: >=3.12
Description-Content-Type: text/markdown

# nucleai

**AI Support for Fusion Experiments**

`nucleai` is a codegen-first library designed to empower fusion research with advanced AI capabilities. It provides a comprehensive workflow for extracting features from fusion data, generating rich descriptions, and creating vector embeddings for both feature images and text.

## Key Features

- **Natural Language Interface**: Ask questions about data and test hypotheses using natural language.
- **Data Search & Retrieval**: Advanced search capabilities using text and feature shapes.
- **Analysis & Display**: Automated analysis and visualization of experimental data.
- **Feature Extraction**: Automated extraction of significant features from experimental fusion data.
- **Vector Embeddings**: Generation of embeddings for feature images and text descriptions to enable semantic understanding.
- **Codegen-First Design**: Built to be driven by AI agents and code generation tools, allowing users to request complex, novel plots and processing pipelines on demand.

## Getting Started

(Instructions for installation and usage will be added here)
