metadata

title: AI Research Assistant
sdk: gradio
sdk_version: 4.38.1
app_file: app.py
license: apache-2.0

🧠 AI Research Assistant

An advanced AI-powered research assistant that combines web search capabilities with contextual awareness to provide comprehensive answers to complex questions.

🌟 Key Features

Real-time Streaming Output: See responses as they're generated for immediate feedback
Contextual Awareness: Incorporates current weather and space weather data
Web Search Integration: Powered by Tavily API for up-to-date information
Smart Caching: Redis-based caching for faster repeated queries
Intelligent Server Monitoring: Clear guidance during model warm-up periods
Accurate Citations: Real sources extracted from search results
Asynchronous Processing: Parallel execution for optimal performance
Responsive Interface: Modern Gradio UI with example queries

🏗️ Architecture

The application follows a modular architecture for maintainability and scalability: myspace134v/ ├── app.py # Main Gradio interface ├── modules/ │ ├── analyzer.py # LLM interaction with streaming │ ├── citation.py # Citation generation and formatting │ ├── context_enhancer.py # Weather and space context (async) │ ├── formatter.py # Response formatting │ ├── input_handler.py # Input validation │ ├── retriever.py # Web search with Tavily │ ├── server_cache.py # Redis caching │ ├── server_monitor.py # Server health monitoring │ ├── status_logger.py # Event logging │ ├── visualizer.py # Output rendering │ └── visualize_uptime.py # System uptime monitoring ├── tests/ # Unit tests ├── requirements.txt # Dependencies └── version.json # Version tracking

🤖 AI Model Information

This assistant uses the DavidAU/OpenAi-GPT-oss-20b-abliterated-uncensored-NEO-Imatrix-gguf model hosted on Hugging Face Endpoints. This is a powerful open-source language model with:

20 Billion Parameters: Capable of handling complex reasoning tasks
Extended Context Window: Supports up to 8192 tokens per response
Uncensored Capabilities: Provides comprehensive answers without artificial limitations
Specialized Training: Optimized for research and analytical tasks

🔧 API Integrations

Service	Purpose	Usage
Tavily	Web Search	Real-time information retrieval
Hugging Face Inference	LLM Processing	Natural language understanding
Redis	Caching	Performance optimization
NASA	Space Data	Astronomical context
OpenWeatherMap	Weather Data	Environmental context

⚡ Enhanced Features

🔁 Streaming Output

Responses stream in real-time, allowing users to start reading before the complete answer is generated. This creates a more natural conversational experience.

📚 Dynamic Citations

All information is properly sourced with clickable links to original content, ensuring transparency and enabling further exploration.

⚡ Asynchronous Operations

Weather data, space weather, and web searches run in parallel, significantly reducing response times.

🧠 Contextual Intelligence

Each query is enhanced with:

Current weather conditions
Recent space events
Accurate timestamps

🛡️ Server State Management

Intelligent monitoring detects when the model server is initializing and provides clear user guidance with estimated wait times.

🚀 Getting Started

Prerequisites

Python 3.8+
Hugging Face account and token
API keys for Tavily, NASA, and OpenWeatherMap
Redis instance for caching

Setup Instructions

Clone the repository

Set up required environment variables:

export HF_TOKEN="your_hugging_face_token"
export TAVILY_API_KEY="your_tavily_api_key"
export REDIS_HOST="your_redis_host"
export REDIS_PORT="your_redis_port"
export REDIS_USERNAME="your_redis_username"
export REDIS_PASSWORD="your_redis_password"
export NASA_API_KEY="your_nasa_api_key"
export OPENWEATHER_API_KEY="your_openweather_api_key"
Install dependencies:

pip install -r requirements.txt Run the application:

python app.py 📊 System Monitoring The assistant includes built-in monitoring capabilities:

Server Health Tracking: Detects and reports server state changes Performance Metrics: Logs request processing times Uptime Monitoring: Tracks system availability Failure Recovery: Automatic handling of transient errors 📋 Example Queries Try these sample questions to see the assistant in action:

"What are the latest developments in fusion energy research?" "How does climate change impact global food security?" "Explain the significance of recent Mars rover discoveries" "What are the economic implications of AI advancement?" 📄 License This project is licensed under the Apache 2.0 License - see the LICENSE file for details.

🤝 Contributing Contributions are welcome! Please feel free to submit a Pull Request.

📞 Support For issues, questions, or feedback, please open an issue on the repository.