Quick Start¶
Get up and running with LORG in 5 minutes.
Prerequisites¶
Make sure you have:
- LORG installed (
npm install -g lorg) - OpenAI API key set (
export OPENAI_API_KEY=your-key)
Your First Search¶
Using the CLI¶
Run a simple search:
You'll see output like:
Searching for: "What is machine learning?"
=== LLM Answer ===
Machine learning is a subset of artificial intelligence that enables
computers to learn and improve from experience without being explicitly
programmed...
=== Knowledge Graph ===
{
"entities": ["machine learning", "artificial intelligence", "algorithms"],
"relationships": [...]
}
=== Keywords ===
machine learning, AI, algorithms, neural networks, data science
=== Search Results ===
1. Introduction to Machine Learning
Score: 0.92
Source: https://example.com/ml-intro
Content: Machine learning is a method of data analysis...
Total tokens used: 1523
Using as a Library¶
import LorgSearch from 'lorg';
async function search() {
const searcher = new LorgSearch(process.env.OPENAI_API_KEY);
const results = await searcher.search('What is machine learning?', {
model: 'gpt-4o-mini',
maxResults: 5
});
console.log('Answer:', results.answer);
console.log('Keywords:', results.keywords);
console.log('Top result:', results.results[0]);
console.log('Tokens used:', results.tokenCount);
}
search();
Starting the Server¶
Start LORG as an API server:
Make a search request:
curl -X POST http://localhost:3000/search \
-H "Content-Type: application/json" \
-d '{"query": "What is machine learning?"}'
Understanding Results¶
LORG returns structured results containing:
| Field | Description |
|---|---|
answer | AI-generated answer to your query |
knowledgeGraph | Structured representation of entities and relationships |
keywords | Extracted search keywords |
results | Array of scored search results with extracted content |
tokenCount | Total OpenAI tokens used |
Next Steps¶
- Configuration - Customize LORG settings
- CLI Usage - Learn all CLI commands
- Library Usage - Integrate into your code