Spatial Speak

Spatial Speak is an AI-powered question-answering system designed to assist scientists in querying and extracting insights from space agriculture-based experiments. Utilizing Retrieval-Augmented Generation (RAG), Spatial Speak enables efficient exploration of complex datasets.

Overview

The primary objective of Spatial Speak is to provide researchers with a tool to navigate and interpret data from spaceflight experiments involving plant biology. The system accesses datasets from NASA's Open Science Data Repository (OSDR), focusing on studies related to plants under spaceflight and high-altitude conditions. These datasets include transcriptional profiles and other relevant biological data from experiments conducted aboard the International Space Station (ISS) and other spaceflight missions.

Features

• Natural Language Processing: Allows users to pose questions in natural language and receive concise, relevant answers based on the underlying datasets.
• Voice Input Capability: Supports voice queries through integration with AssemblyAI's speech-to-text API, enabling hands-free interaction.
• Session-Based Conversation Tracking: Maintains context across multiple queries within a session for a coherent conversational experience.
• Chat History Management: Provides a record of previous interactions for reference and further analysis.

Setup

1. Clone the Repository:

   git clone <repository-url>
   cd spatial_speak

2. Install Dependencies:

   pip install -r requirements.txt

3. Configure API Keys:

   • Open configure.py.
   • Insert your AssemblyAI API key to enable voice input functionality.

4. Run the Application:

   streamlit run app.py

Project Structure

• app.py: Main application script utilizing Streamlit for the user interface.
• stt.py: Handles speech-to-text conversion using AssemblyAI's API.
• chatbot_model.py: Contains functions and classes related to LLM integration and response generation.
• demo_notebook.py: Jupyter notebook demonstrating model utilities and sample interactions.
• configure.py: Stores configuration settings, including API keys.
• llama_stuff/: Directory containing reference data and resources for the LLM model.

Data Source

Spatial Speak utilizes datasets from NASA's Open Science Data Repository (OSDR), specifically focusing on experiments involving plant root seedlings under spaceflight and high-altitude conditions. These studies provide valuable insights into plant biology in microgravity environments, contributing to the advancement of space agriculture research.