README.md

metadata

title: Agentic AI Patterns for Scientific Q&A
emoji: 🤖
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 4.44.0
app_file: app.py
pinned: false
license: mit

🤖 Agentic AI Patterns for Scientific Q&A

A comprehensive comparison of 6 different AI reasoning approaches for answering scientific questions, powered by OpenAI GPT-4o-mini and OpenAlex scientific database.

🎯 Features

6 AI Reasoning Patterns:

1. 🔤 Pure LLM - Direct response without enhancement
2. 🪞 Reflection - Agent reviews and improves its own answer
3. 📋 Planning - Creates and follows a systematic research plan
4. 🔧 Tool Use - Uses OpenAlex to find relevant scientific papers
5. 👥 Multi-Agent - Multiple specialized agents collaborate
6. 🧠 Chain of Thought - Step-by-step logical reasoning

⚖️ Comparison Mode

Run all 6 patterns simultaneously to compare:

• Processing speed
• Confidence levels
• Source utilization
• Answer quality

🔬 Scientific Integration

• OpenAlex Database: Access to 240M+ scientific papers
• Real-time Literature Search: Find relevant research for any question
• Citation Analysis: Get papers with citation counts and metadata
• Multi-domain Support: Physics, Biology, Chemistry, Medicine, Computer Science, and more

🎨 Beautiful Interface

• Modern Design: Glass-morphism effects with gradient backgrounds
• Dark/Light Theme: Automatic theme adaptation
• Responsive Layout: Works on desktop and mobile
• Interactive Elements: Hover effects and smooth transitions
• Example Questions: Pre-loaded scientific questions to try

🚀 How to Use

1. Get OpenAI API Key: Visit OpenAI Platform to get your API key
2. Enter your question in the text area (or use example questions)
3. Provide your OpenAI API key in the API key field
4. Choose a pattern to see how different AI approaches handle your question
5. Compare all patterns to see which works best for your question type

🧪 Example Questions

• "What are the main mechanisms behind CRISPR-Cas9 gene editing and what are its current limitations?"
• "How does climate change affect ocean acidification and marine ecosystems?"
• "What is the relationship between gut microbiome diversity and human immune system function?"
• "Explain the current understanding of dark matter and dark energy in cosmology."
• "What are the key differences between mRNA vaccines and traditional vaccines?"

🛠️ Technical Details

• Framework: Gradio 4.44.0
• AI Model: OpenAI GPT-4o-mini
• Database: OpenAlex API
• Language: Python with async/await
• Styling: Custom CSS with theme support

📊 Pattern Comparison

Each pattern offers unique advantages:

• Pure LLM: Fastest, baseline performance
• Reflection: Higher accuracy through self-correction
• Planning: Systematic approach for complex questions
• Tool Use: Evidence-based answers with citations
• Multi-Agent: Multiple perspectives and expertise
• Chain of Thought: Transparent reasoning process

🔒 Privacy & Security

• User-provided API keys: Users must bring their own OpenAI API keys
• No API key storage: Keys are only used during the session
• No data logging: No conversation data is stored
• Ephemeral processing: All data is processed in real-time only
• Open source: Fully transparent code

🎓 Educational Value

Perfect for:

• Researchers comparing AI reasoning approaches
• Students learning about agentic AI patterns
• Educators demonstrating different AI capabilities
• Developers understanding advanced prompt engineering

🌟 Based on Research

Implements proven agentic AI patterns from recent research in AI reasoning and tool use, adapted specifically for scientific question answering.

---

Experiment with different AI reasoning patterns and discover which approach works best for your scientific questions!