W
WazBotDocs

Introduction to WazBot

WazBot is an AI-powered platform that lets you build, deploy, and manage intelligent chatbots on WhatsApp Business. Whether you're a small business owner looking to automate customer support or an enterprise managing thousands of conversations daily, WazBot scales with you.

New to WazBot?
Head to the Getting Started guide to have your first chatbot live in under 5 minutes.

What Can WazBot Do?

Smart Conversations

Natural language understanding powered by the latest LLMs — GPT-4o, Claude 3.5, and more.

AI Agent System

Custom AI agents with personality, knowledge base, and tool-calling capabilities.

Analytics Dashboard

Real-time insights into conversation volume, satisfaction scores, and agent performance.

Third-Party Integrations

Connect to CRMs, payment gateways, calendars, and hundreds of other services.

How It Works

  1. Connect — Link your WhatsApp Business account via the official Cloud API.
  2. Configure — Set up an AI agent with a custom personality, knowledge base, and tools.
  3. Deploy — Go live instantly. Your bot begins handling conversations in real time.
  4. Optimize — Use analytics and conversation logs to improve responses over time.

Core Concepts

Agents

An Agent is the AI brain behind your chatbot. Each agent has its own system prompt, knowledge base, tool integrations, and behavioral settings. You can create multiple agents for different use cases (e.g., Sales Bot, Support Bot, Booking Bot).

Knowledge Base

Upload documents (PDFs, CSVs, web pages) that your agent can reference when answering questions. WazBot uses RAG (Retrieval-Augmented Generation) to pull relevant context from your knowledge base and include it in responses.

Conversations

Every WhatsApp chat thread is a Conversation. WazBot maintains context within a conversation window (configurable from 1 hour to 30 days), so returning customers pick up right where they left off.

Webhooks & API

Use the REST API to programmatically manage agents, send messages, and receive real-time event webhooks for new messages, conversation status changes, and more.

Architecture Overview

WazBot sits between the WhatsApp Cloud API and your AI models. Messages flow through a processing pipeline:

  1. Ingestion — Incoming message received via WhatsApp webhook.
  2. Context Retrieval — Conversation history and knowledge base documents fetched.
  3. AI Processing — LLM generates a response based on agent config + context.
  4. Tool Execution — If the AI decides to call a tool (API, database, etc.), it executes and uses the result.
  5. Delivery — Response sent back to the customer via WhatsApp.
Latency
Average end-to-end response time is under 3 seconds for standard text replies.

Next Steps

Next
Getting Started