Advanced Memory System
Built-in vector memory with semantic search, conversation history, and intelligent context retrieval using PostgreSQL + pgvector.
The most powerful TypeScript SDK for creating intelligent AI agents with advanced memory capabilities, tool integration, and seamless chat experiences.

import { Aisar } from '@aisar/sdk';
// Initialize the SDK
const aisar = new Aisar({
apiKey: 'your-api-key',
baseUrl: 'https://api.aisar.ai'
});
// Create an agent with memory
const agent = await aisar.createAgent({
name: 'Support Assistant',
instructions: 'You are a helpful customer support agent.',
model: 'gpt-4',
memoryEnabled: true
});
// Start a conversation with persistent memory
const session = await aisar.createSession(agent.id);
const response = await session.addMessage('user', 'Hello! I need help with my order.');
console.log(response.content); // AI response with contextUnlike basic chatbots, Aisar agents remember and learn from conversations using advanced vector embeddings and semantic search.
Agents can execute custom functions, query databases, call APIs, and integrate with your existing systems seamlessly.
Built for production with comprehensive error handling, monitoring, analytics, and scalable architecture.
Beautiful TypeScript types, excellent documentation, interactive examples, and a thriving community.
"Aisar SDK transformed how we build AI agents. The memory system is incredibly powerful and the developer experience is fantastic."
— Sarah Chen, CTO at TechCorp
"The TypeScript support and documentation are top-notch. We went from prototype to production in days, not weeks."
— Michael Rodriguez, Senior Developer
Just getting started?
The Getting Started Guide provides a comprehensive introduction to building your first AI agent.
Looking for specific functionality?
Check out our API Reference for detailed documentation of all classes and methods.
Need real-world examples?
Explore our Examples Gallery with ready-to-run code for common use cases.