WORK IN PROGRESS!
RADIAN is an interactive data analysis platform that combines Large Language Models with a user-friendly R Shiny interface. It enables researchers to conduct data analysis through a combination of natural conversation with an AI agent and manual code execution, allowing for a more flexible and interactive analytical workflow.
- Interactive Analysis Sessions: Conduct data analysis with real-time AI assistance
- Intelligent AI Agent: Communicates with users, suggests code, and can execute analysis directly
- Context-Aware: Maintains full history of code execution, results, and conversation
- REPL Environment: Notebook-style interface
- Data Management: Import and analyze data from various formats (CSV, Excel, etc.)
- Environment Tracking: Monitor variables, datasets, and analysis state in real-time
RADIAN's main interface showing the chat and code execution environment
Example of an interactive data analysis session with AI assistance
- Clone the repository
git clone https://github.com/demirev/radian.gitandcd radian - Create a
.envfile with the required variables, including LLM provider API keys (seedocker-compose.ymlfor the required variables) - Run
docker compose up -d
For detailed information about the project architecture and features, a rudimentary Project Documentation is available.
- Georgi Demirev (first_last[et]proton[dot]me).