💬 Language AI

Build Your Chatbot

Design intents, add training phrases, write responses — then chat with the bot you built! See exactly how real chatbots like Siri and Alexa work inside.

🧩 Design Intents
📝 Add Training Phrases
💬 Chat & Test
📊 Confusion Matrix
🏆 Badge

How Chatbots Understand You

🎯

Intents

An intent is the "goal" behind a message. "What's the weather?" and "Will it rain?" are the same intent: asking about weather.

📚

Training Phrases

Many ways of saying the same thing. The more phrases you add, the smarter your bot becomes at recognising user messages.

🔍

Intent Matching

The bot compares your message to all training phrases using keyword similarity and picks the best matching intent.

📊

Confusion Matrix

A grid showing which intents the bot gets right vs. confuses. The diagonal should be bright green for a good bot!

💬
Wizzy the AI Tutor
Welcome to the chatbot lab! 🎉 First, let's design the intents — the things your bot will understand. I've added 3 starter intents for a School Helper Bot. Add more, or create your own theme — a weather bot, a food bot, whatever you like!

Step 1 — Design Your Intents

3 intents

➕ Create New Intent

👈 Click an intent to see details
💬
Wizzy the AI Tutor
Now teach your bot! For each intent, add at least 3 training phrases — different ways a user might say the same thing. Also write 2+ responses the bot will randomly pick when it recognises that intent. More variety = smarter bot!

Step 2 — Train Each Intent

0 / 0 ready
👈 Select an intent to train it
💬
Wizzy the AI Tutor
Your bot is alive! 🤖 Type messages and see if it understands you. Watch the intent match bars on the right — they show how confident the bot is about each intent. Try to confuse it! Can you find messages it gets wrong?

Chat with Your Bot

Testing...
Intent Match Scores
Send a message to see scores

📊 Chat Stats

Messages sent0
Understood0
Not understood0
Accuracy
💬
Wizzy the AI Tutor
This Confusion Matrix shows which intents your bot confuses with each other. Green diagonal = correct! Red off-diagonal = confused. Add more training phrases to fix confused pairs. Real AI engineers use this exact tool every day!

Confusion Matrix Analysis

💬
Wizzy the AI Tutor
🎊 Amazing! You just built a real intent-based chatbot — the same architecture used by Google Assistant, Amazon Alexa, and Apple Siri! You understand intents, training phrases, confidence scores, and confusion matrices. That's real NLP engineering!
💬

Chatbot Builder Badge!

You designed intents, trained your bot, and analysed its confusion matrix!

💬 WhizzStep AI Lab
This certifies that
Student Name
has built an Intent-Based Chatbot from scratch
NLP Engineer
Intent Designer
Bot Builder
Built — intents · whizzstep.in

Key Concepts Mastered

Intent

🎯 The Goal

The meaning behind a user message. "Book a flight" and "I want to fly to Delhi" are the same intent.

Training Data

📚 Teaching Examples

The set of example phrases that teach the bot what each intent looks like. More = better.

Confidence Score

📊 How Sure?

A number between 0–1 showing how confident the bot is about its chosen intent. Below 0.3 = confused.

Confusion Matrix

🔢 Error Analysis

A table showing correct vs. incorrect predictions. Engineers use this to find which intents need more training data.

Fallback

🤷 I Don't Know

When confidence is too low, the bot uses a fallback response: "Sorry, I didn't understand that."

NLP

🧠 Natural Language

Natural Language Processing — the field of AI that helps computers understand human language.