Examples
Learn by example with real-world use cases and complete code samples.
Chat
Customer Support Bot
Build an intelligent support bot with memory and reasoning
Search
Semantic Search Engine
Create a powerful search with natural language understanding
RAG
Document Q&A System
Answer questions from your documentation automatically
Analytics
Sales Intelligence
Analyze deals and provide actionable insights
ML
E-commerce Recommendations
Personalized product recommendations with reasoning
Dev Tools
Code Assistant
Build a coding copilot with context awareness
Featured: Customer Support Bot
Build an intelligent support bot that remembers past conversations and reasons about customer issues:
What You'll Learn
This example demonstrates Memory Layer for context storage, Cognition Layer for reasoning, and how to build context-aware AI applications.
support-bot.tstypescript
import { CognitiveXClient } from '@cognitivex/sdk';
const client = new CognitiveXClient({
apiKey: process.env.COGNITIVEX_API_KEY
});
async function handleCustomerQuery(
customerId: string,
query: string
) {
// Store the query in memory
await client.memory.store({
content: query,
tags: ['customer-query', customerId],
metadata: {
customerId,
timestamp: new Date().toISOString()
}
});
// Search for relevant past interactions
const pastContext = await client.memory.search({
query,
tags: [customerId],
limit: 5
});
// Reason with full context
const response = await client.cognition.reason({
query,
context: pastContext.map(c => c.id),
reflection: true,
systemPrompt: `You are a helpful customer support agent.
Use past conversations to provide personalized help.`
});
return {
answer: response.answer,
reasoning: response.reasoning,
sources: pastContext
};
}
// Example usage
const result = await handleCustomerQuery(
'user_123',
'My order hasn\'t arrived yet'
);
console.log('Response:', result.answer);
console.log('Reasoning:', result.reasoning);Cost Savings
With semantic caching enabled, similar customer queries cost 85% less. The first query costs ~$0.05, but subsequent similar queries are served from cache at no cost!
Quick Examples
Semantic Search
typescript
// Index documents
const docs = [
"CognitiveX reduces AI costs by 85%",
"Multi-LLM support for OpenAI, Anthropic, Google",
"Enterprise-grade security with SOC2 compliance"
];
for (const doc of docs) {
await client.memory.store({ content: doc });
}
// Search with natural language
const results = await client.memory.search({
query: "cost savings",
limit: 3
});
results.forEach(r => {
console.log(r.content, '- Score:', r.score);
});Multi-Step Reasoning
typescript
// Complex reasoning with reflection
const analysis = await client.cognition.reason({
query: `Analyze Q4 sales data and recommend
strategies for Q1 growth`,
reflection: true,
temperature: 0.7
});
console.log('Analysis:', analysis.answer);
console.log('Thought process:', analysis.reasoning);
console.log('Confidence:', analysis.confidence);Use Case Library
Customer Support
- • Contextual chatbot
- • Ticket classification
- • Sentiment analysis
Sales & Marketing
- • Lead qualification
- • Content generation
- • Personalization
Development
- • Code review assistant
- • Documentation generator
- • Bug triage
Operations
- • Log analysis
- • Incident response
- • Root cause analysis