🤖 Module 6: AI Analyzer

Chat with two AI systems about ADIZ spikes: Global LLM (GPT-4o-mini with external knowledge) and Local RAG (conversational AI trained only on your dataset)

Two Complementary Approaches:
🌐 Global LLM (GPT-4o-mini): Full geopolitical knowledge + your data. Best for "why did this happen?" with external context.
🔍 Local RAG: Conversational AI that uses ONLY your dataset. Semantic search + natural language answers. Best for dataset exploration and fact-checking.
← Dashboard Visualizer Module 1 Module 2 Module 3
🌐

Global LLM (GPT-4o-mini)

External knowledge + your dataset

Ready
Model: GPT-4o-mini
Knowledge: Taiwan-China relations, US policy, regional geopolitics + your ADIZ data
Best for: "Why did X happen?", external context, predictions
👋 Hi! I'm GPT-4o-mini with access to your ADIZ dataset AND broader geopolitical knowledge. Ask me anything like: • "Why were there 38 incursions on October 15, 2024?" • "How does this compare to past Taiwan elections?" • "What might China do if Taiwan gets more F-16s?" I combine your data with world events and context.

💡 Example Questions:

Why were there 38 incursions on October 15, 2024?
How does China's behavior in 2024 compare to 2022?
What would likely happen if the US sold F-35s to Taiwan?
🔍

Local RAG

Conversational AI - dataset only

Loading...
Status: Loading models...
Knowledge: ONLY your ADIZ dataset (no external info)
Best for: Exploring data, fact-checking, finding patterns
⏳ Loading RAG system... This takes ~1 min first time. Once ready, I'm a conversational AI trained ONLY on your ADIZ data. I can: • Answer open-ended questions about patterns • Have back-and-forth conversations • Explore your dataset with you • But I know NOTHING outside your data!