Skip to content

Text & Voice Chatbot Features

Products: Text Chatbot + Voice Chatbot
Type: Core offerings (text standard, voice premium)
Target Users: All customer segments
Business Value: Widest adoption, easiest deployment

🎯 Overview

Text and Voice Chatbots provide fast, intelligent conversations without the complexity of 3D avatars. Perfect for customers who need quick deployment, minimal resources, or prefer simpler interfaces.

Text Chatbot: Clean chat interface with instant AI responses
Voice Chatbot: Speech-enabled conversations with audio playback

💬 Text Chatbot Features

1. Clean Chat Interface

What it does:
Simple, familiar chat interface that users know how to use instantly.

Key Features:

  • Clean, minimal design
  • Instant message delivery
  • Typing indicators
  • Auto-scroll to latest message
  • Message timestamps
  • Session management
  • Mobile responsive

User Benefit:
Zero learning curve. Users start chatting immediately without tutorials.

Technical Foundation:

  • Frontend: React chat component (225 lines)
  • Backend: response-text-chatbot-service (Port 8012)
  • Storage: MongoDB chatbot_history

2. Lightning-Fast Responses

What it does:
Optimized for speed - responses in under 2 seconds.

How It's Faster Than 3D:

  • No TTS generation (saves ~1.2s)
  • No lip-sync processing (saves ~0.7s)
  • Simpler rendering (saves ~0.5s)
  • Result: 2-3s avg vs 4-8s for 3D

Response Time Breakdown:

Vector search:      ~120ms
LLM generation:   ~1,500ms
Markdown cleanup:   ~50ms
Save to DB:        ~100ms
─────────────────────────
Total:           ~1,770ms

User Benefit:
Feels instant. Higher user satisfaction for quick questions.

3. Same Intelligent AI

What it does:
Identical RAG pipeline and AI capabilities as 3D chatbot.

AI Features:

  • Same 11 LLM models available
  • Same vector search (Milvus)
  • Same knowledge base options
  • Same context memory
  • Same accuracy (90%+)

Data Sources:

  • Website crawling: 25 URLs (vs 50 for 3D)
  • PDF uploads: Multiple files
  • Manual text input
  • Q&A pairs

User Benefit:
Premium AI intelligence without premium price.

4. Simple Customization

Configuration Options:

  • Purpose selection (Sales, Support, FAQ, etc.)
  • System prompts
  • Greeting message
  • Brand colors
  • Custom CSS (advanced)

No Avatar/Voice:

  • Simpler setup process
  • Faster to deploy
  • Lower resource usage

5. Easy Deployment

Widget SDK:

<script
  src="https://machineavatars.com/sdk/text-chat.js"
  data-project="[id]"
></script>

iFrame:

<iframe
  src="https://app.machineavatars.com/text-chatbot/[id]"
  width="350"
  height="500"
></iframe>

User Benefit:
Add to website in under 5 minutes.

🎤 Voice Chatbot Features

1. Speech-Enabled Conversations

What it does:
Talk to your chatbot like a virtual assistant (Siri/Alexa style).

Key Features:

  • Speech Recognition: Web Speech API (browser-based)
  • Voice Synthesis: Azure TTS (10 premium voices)
  • Audio Playback: Base64-encoded streaming
  • Dual Input: Speak OR type

User Benefit:
Hands-free operation. Perfect for accessibility, mobile users, or while multitasking.

Technical Foundation:

  • Frontend: voice-chatbot/page.tsx (384 lines)
  • Backend: response-voice-chatbot-service (Port 8013)
  • Browser API: Web Speech API (free, no server cost)

2. Natural Voice Responses

What it does:
AI responses are spoken aloud in natural voice.

Voice Options: (Same 10 as 3D)

  1. Esther Howard (Female)
  2. Erica Rose (Female 2)
  3. Sophia Martinez (Female 3)
  4. Rose Bennett (Female 4)
  5. Liam Johnson (Male)
  6. Adrian Foster (Male 2)
  7. Benjamin Hayes (Male 3)
  8. Caleb Reed (Male 4)
  9. Emily Parker (Female 5)
  10. Female 6

Audio Format:

  • Azure Neural TTS
  • Base64 audio delivery
  • Auto-play on response
  • Volume control
  • Pause/resume

User Benefit:
Natural conversations without reading. Great for accessibility (vision impaired).

3. Hybrid Input/Output

What it does:
Mix speech and text however you prefer.

Flexible Modes:

  • Speak + Listen: Full voice conversation
  • Speak + Read: Voice input, text output
  • Type + Listen: Text input, voice output
  • Type + Read: Full text (like text chatbot)

User Benefit:
Adapt to environment. Voice in car, text in office.

4. Same Data Training

What it does:
Identical knowledge base options as 3D and Text.

Data Sources:

  • Website: 50 URLs (same as 3D!)
  • Files: PDF, DOCX
  • Text: Manual input
  • Q&A: Predefined pairs
  • Voice Avatar: Select voice personality

Extra Step: Voice selection (adds 1 wizard step vs text)

5. Mobile-Optimized

What it does:
Works great on phones and tablets.

Mobile Features:

  • Touch-optimized buttons
  • Large microphone button
  • Auto-permissions for mic
  • Responsive audio player
  • Works offline (speech recognition)

User Benefit:
Perfect for on-the-go users.

📊 Feature Comparison: Text vs Voice vs 3D

Feature Text Voice 3D
Visual Chat bubbles Chat bubbles 3D Avatar
Audio ❌ ✅ TTS ✅ TTS + Lip-Sync
Input Text only Speech + Text Text only*
Response Time ~2s ~3s ~5s
Setup Time 5 min 10 min 15 min
Website URLs 25 50 50
File Upload ✅ ✅ ✅
LLM Models All 11 All 11 All 11
Mobile Support ✅ ✅ Limited
Bandwidth Low Medium High
Best For FAQ, Simple Accessibility, Mobile Premium, Engagement
Price $ $$ $$$

*3D voice input in roadmap

💼 Use Cases

Text Chatbot Best For:

1. Customer Support FAQ

  • Fast answers
  • Many concurrent users
  • Low bandwidth environments

2. Internal Knowledge Base

  • Employee self-service
  • Documentation Q&A
  • Quick lookups

3. Simple Lead Gen

  • Basic qualification
  • Contact collection
  • Appointment booking

Voice Chatbot Best For:

1. Accessibility

  • Vision-impaired users
  • Hands-free operation
  • Reading disabilities

2. Mobile-First

  • Drivers (hands-free)
  • Warehouse workers
  • Field technicians

3. Elderly/Non-Tech Users

  • Speaking easier than typing
  • More natural interaction
  • Better engagement

🔧 Configuration Guide

Text Chatbot Setup (5 Steps)

  1. Create Chatbot

  2. Select "Text Chatbot"

  3. Name your chatbot

  4. Choose purpose (Sales/Support/etc.)

  5. Add Data Sources

  6. Enter website URL (crawls up to 25 pages)

  7. OR upload PDF/DOCX
  8. OR paste text content

  9. OR create Q&A pairs

  10. Customize Appearance

  11. Set greeting message

  12. Choose brand colors

  13. Configure bubble position

  14. Test

  15. Try sample questions

  16. Verify responses

  17. Check knowledge coverage

  18. Deploy

  19. Copy widget code
  20. Paste in website
  21. Done!

Time: ~5 minutes

Voice Chatbot Setup (6 Steps)

Same as text PLUS:

Step 4.5: Select Voice

  • Choose from 10 Azure TTS voices
  • Preview each voice
  • Select gender and tone
  • Test voice with sample text

Time: ~10 minutes

📈 Performance Specifications

Text Chatbot

Response Time: 1.5-2.5 seconds average

Breakdown:

  • Milvus search: 100-150ms
  • LLM generation: 1,200-1,800ms
  • Processing: 100-200ms

Concurrent Users: 5,000+

Accuracy: 92% (same as 3D/Voice)

Uptime: 99.5% SLA

Voice Chatbot

Response Time: 2.5-3.5 seconds average

Breakdown:

  • Milvus search: 100-150ms
  • LLM generation: 1,200-1,800ms
  • TTS synthesis: 1,000-1,300ms
  • Processing: 100-200ms

Concurrent Users: 3,000+

Accuracy: 92% (text), 85% (speech recognition in noisy environments)

Uptime: 99.5% SLA

🔗 Technical Integration

Text Chatbot

Frontend: text-chatbot/page.tsx (225 lines)
Backend: response-text-chatbot-service (Port 8012)
Wizard: chatbot-data-source/page.tsx (604 lines)

Collections:

  • chatbot_history
  • chatbot_selections
  • files

APIs:

  • POST /v2/get-response-text-chatbot
  • Streaming variant available

Voice Chatbot

Frontend: voice-chatbot/page.tsx (384 lines)
Backend: response-voice-chatbot-service (Port 8013)
Wizard: voice-chatbot-data-source/page.tsx (656 lines)

Collections:

  • chatbot_history
  • chatbot_selections
  • files
  • voice_selections

APIs:

  • POST /v2/get-response-voice-chatbot
  • Returns: {text, audio_base64}

Browser APIs:

  • Web Speech API (recognition)
  • Audio element (playback)

🚨 Limitations

Text Chatbot

  • No voice/audio
  • No visual avatar
  • Text-only responses

Workarounds:

  • Use voice chatbot if audio needed
  • Upgrade to 3D for visual engagement

Voice Chatbot

  • Speech recognition accuracy varies by browser
  • Requires microphone permission
  • Background noise affects accuracy
  • No lip-sync (audio only)

Workarounds:

  • Provide fallback text input always
  • Chrome recommended (best speech API)
  • Upgrade to 3D for lip-sync

Backend:

Frontend:

Compare:

"Simple, fast, intelligent - choose your conversation style." 💬🎤