> ## Documentation Index
> Fetch the complete documentation index at: https://docs.wapikit.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Conversation Engine

> Learn how Wapikit autonomously replies to customers with human-like intelligence and full context awareness

Wapikit’s **Conversation Engine** is not your average chatbot. It’s an AI-powered operational brain that handles your **customer conversations in real-time**, just like a human teammate—but better.

<Note>
  Wapikit replies to **every customer message in under 1 minute** with context-aware, emotionally intelligent responses.
</Note>

***

## What It Does

* 📥 **Handles all incoming WhatsApp messages 24/7**
* 🧠 **Understands customer intent**, not just keywords
* 🛍️ **Recommends products**, answers FAQs, provides policy info
* 🧾 **Handles orders**, from status updates to changes and cancellations
* 🗂️ **Maintains full customer memory** across sessions
* 🎯 **Personalizes responses** based on user profile, behavior, and past interactions
* 🤖 **Works even while you're sleeping**

***

## How It Works

Wapikit’s AI engine listens to every incoming message and does three things instantly:

### 1. Intent Detection

Understands whether the user is asking a question, requesting help, trying to purchase, or just exploring.

> “Can I get this in a different size?”\
> → Detected as: *Product Variant Inquiry*

### 2. Context Recall

Wapikit accesses customer context, including:

* Last campaign message received
* Past orders and interactions
* Tagged preferences or behavior
* CRM attributes (location, order value, loyalty status)

### 3. Intelligent Action

Once intent is detected and context is loaded, Wapikit replies or executes actions:

* Replies to order-related queries using live order data
* Recommends the most relevant product
* Pushes follow-up campaigns if needed
* Escalates to human when necessary

***

## Powered by Knowledge Base + Memory

The engine is fueled by your **custom Knowledge Base**, where you define your:

* FAQs
* Return/refund/cancellation policies
* Delivery timelines
* Product info and attributes

<Note>
  Wapikit **auto-maps knowledge base answers to customer intents**, and learns which ones perform best.
</Note>

And because it **remembers each customer’s profile** and updates it over time, replies get smarter and more relevant with every message.

***

## Designed for D2C & Service Brands

Whether you’re a clothing brand handling size and style queries, or a service provider managing bookings and support:

* ⚡ Wapikit scales your operations without hiring
* 👂 Gives every customer instant support that feels personal
* 📈 Boosts conversions, CSAT, and retention

***

## Real-World Example

> **Customer**: "Hey, when will my order #1234 arrive?"

Wapikit sees:

* This is an order status inquiry
* Customer ordered a ₹1,299 Kurta Set last Thursday
* Standard shipping = 5 days

Wapikit replies:

> “Hi Priya 👋 Your order placed on May 24th is expected to arrive by May 29th. We’ll send you a tracking link soon. Let me know if you need anything else!”

***

## Need Full Control?

You can customize:

* Tone of voice (friendly, formal, humorous)
* What to escalate to humans
* Product recommendation logic
* What data to store in memory

***

## Learn More

Want to configure your own knowledge base and AI behavior?

👉 [See Knowledge Base Setup](/guide/conversation-engine/knowledge-base)

***

## Built for Scale

* Handles thousands of messages simultaneously
* Optimized for fast response `<1 min`
* Learns over time using performance feedback loops

***

## Summary

Wapikit’s Conversation Engine is your always-on, emotionally aware, operational team member—handling sales, support, and everything in between with `<1 min response time`, 24/7.

***

Need help fine-tuning your engine? [Give this a read](/knowledge-base)