Build an AI Chatbot for Odoo Without Enterprise

Customer → Website chat widget → Webhook → n8n
                                              │
                                              ├─→ OpenAI: understand intent
                                              │
                                              ├─→ Odoo API: fetch relevant data
                                              │     (products, orders, invoices,
                                              │      knowledge base articles)
                                              │
                                              ├─→ OpenAI: generate response
                                              │     with Odoo data as context
                                              │
                                              └─→ Webhook response → Customer

services:
  # ... your existing odoo and db services ...

  n8n:
    image: n8nio/n8n:latest
    restart: unless-stopped
    ports:
      - "5678:5678"
    environment:
      - N8N_BASIC_AUTH_ACTIVE=true
      - N8N_BASIC_AUTH_USER=admin
      - N8N_BASIC_AUTH_PASSWORD=${N8N_PASSWORD:-changethis}
      - N8N_HOST=${N8N_HOST:-n8n.yourdomain.com}
      - N8N_PROTOCOL=https
      - WEBHOOK_URL=https://${N8N_HOST:-n8n.yourdomain.com}/
      - GENERIC_TIMEZONE=UTC
    volumes:
      - n8n_data:/home/node/.n8n
    networks:
      - odoo-network

volumes:
  n8n_data:

docker compose up -d n8n

{
  "message": "What's the status of my order SO-2024-0142?",
  "session_id": "abc123",
  "customer_email": "jane@example.com"
}

// Simple in-memory conversation store
// For production, use Redis or a database
const sessionId = $input.first().json.session_id;
const message = $input.first().json.message;

// Get existing conversation from workflow static data
const staticData = $getWorkflowStaticData('global');
if (!staticData.conversations) {
  staticData.conversations = {};
}
if (!staticData.conversations[sessionId]) {
  staticData.conversations[sessionId] = [];
}

// Add current message
staticData.conversations[sessionId].push({
  role: 'user',
  content: message,
  timestamp: new Date().toISOString()
});

// Keep last 10 messages for context
const history = staticData.conversations[sessionId].slice(-10);

return [{
  json: {
    ...$input.first().json,
    conversation_history: history
  }
}];

You are an intent classifier for an e-commerce support chatbot. Given a customer message, classify it into exactly one category:

- ORDER_STATUS: customer is asking about an order, delivery, or shipment
- PRODUCT_INQUIRY: customer is asking about a product, pricing, or availability
- ACCOUNT_QUESTION: customer is asking about their account, invoices, or payments
- SUPPORT_REQUEST: customer needs help with a problem or wants to file a complaint
- GENERAL: anything else (greetings, small talk, questions about the company)

Respond with ONLY the category name, nothing else.

You are a helpful customer support assistant for [Your Company Name],
an online store running on Odoo. You are friendly, professional, and concise.

RULES:
1. Answer based ONLY on the Odoo data provided in the context. Never make up
   order statuses, prices, or product details.
2. If the data doesn't contain enough information to answer, say so honestly
   and offer to connect the customer with a human agent.
3. Keep responses under 150 words. Customers are chatting, not reading essays.
4. Format prices as currency with 2 decimal places.
5. For order status, translate Odoo states: "sale" = "Confirmed",
   "done" = "Completed", "cancel" = "Cancelled", "draft" = "Draft/Pending".
6. Never reveal internal system details, field names, or technical errors.
   If something goes wrong, say "Let me connect you with a team member."
7. If asked about returns, exchanges, or complaints, always offer to
   create a support ticket.

COMPANY INFO:
- Name: [Your Company Name]
- Return policy: [Your return policy summary]
- Business hours: [Your hours]
- Support email: [Your support email]

{
  "response": "Your order SO-2024-0142 was confirmed on January 15th for $2,450.00. It's currently being prepared for shipment. Based on the expected delivery date, you should receive it by January 22nd. Would you like me to check anything else?",
  "session_id": "abc123",
  "needs_human": false
}

PRODUCT CATALOG:
- Widget Pro ($49.99) - Professional-grade widget, stainless steel
- Widget Basic ($19.99) - Entry-level widget, plastic housing
- Widget Kit ($79.99) - Pro + Basic + accessories, best value
- Spare Parts Pack ($12.99) - Replacement parts for all models

PRICING RULES:
- Orders over $500 qualify for free shipping
- Wholesale pricing available for orders of 50+ units (direct to sales team)
- Active promotion: 15% off Widget Pro through March 31, 2026 (code: SPRING15)

ESCALATION RULES:
- Any mention of "refund" or "cancel" → offer to create a support ticket
- Questions about custom orders → transfer to human agent
- Complaints about quality → apologize, create ticket, flag as priority

<!-- Add before </body> -->
<div id="oec-chat-widget">
  <div id="oec-chat-toggle" onclick="toggleChat()">
    <svg width="24" height="24" viewBox="0 0 24 24" fill="white">
      <path d="M20 2H4c-1.1 0-2 .9-2 2v18l4-4h14c1.1 0 2-.9 2-2V4c0-1.1-.9-2-2-2z"/>
    </svg>
  </div>
  <div id="oec-chat-window" style="display:none;">
    <div id="oec-chat-header">
      <strong>Support Chat</strong>
      <span onclick="toggleChat()" style="cursor:pointer">&times;</span>
    </div>
    <div id="oec-chat-messages"></div>
    <div id="oec-chat-input">
      <input type="text" id="oec-msg" placeholder="Type a message..."
             onkeypress="if(event.key==='Enter')sendMessage()">
      <button onclick="sendMessage()">Send</button>
    </div>
  </div>
</div>

<script>
const WEBHOOK_URL = 'https://n8n.yourdomain.com/webhook/odoo-chatbot';
const SESSION_ID = 'sess_' + Math.random().toString(36).substr(2, 9);

function toggleChat() {
  const win = document.getElementById('oec-chat-window');
  win.style.display = win.style.display === 'none' ? 'flex' : 'none';
}

async function sendMessage() {
  const input = document.getElementById('oec-msg');
  const message = input.value.trim();
  if (!message) return;

  appendMessage('You', message);
  input.value = '';

  try {
    const res = await fetch(WEBHOOK_URL, {
      method: 'POST',
      headers: { 'Content-Type': 'application/json' },
      body: JSON.stringify({
        message: message,
        session_id: SESSION_ID,
        customer_email: getCustomerEmail()
      })
    });
    const data = await res.json();
    appendMessage('Support', data.response);

    if (data.needs_human) {
      appendMessage('System', 'Connecting you with a team member...');
    }
  } catch (err) {
    appendMessage('System', 'Connection issue. Please try again.');
  }
}

function appendMessage(sender, text) {
  const container = document.getElementById('oec-chat-messages');
  const div = document.createElement('div');
  div.className = 'chat-msg ' + (sender === 'You' ? 'user' : 'bot');
  div.innerHTML = '<strong>' + sender + ':</strong> ' + text;
  container.appendChild(div);
  container.scrollTop = container.scrollHeight;
}

function getCustomerEmail() {
  return null; // Return logged-in user's email if available
}
</script>

<style>
#oec-chat-widget {
  position: fixed; bottom: 20px; right: 20px;
  z-index: 9999; font-family: sans-serif;
}
#oec-chat-toggle {
  width: 56px; height: 56px; border-radius: 50%;
  background: #714B67; display: flex; align-items: center;
  justify-content: center; cursor: pointer;
  box-shadow: 0 2px 12px rgba(0,0,0,0.15);
}
#oec-chat-window {
  flex-direction: column; width: 360px; height: 480px;
  background: white; border-radius: 12px;
  box-shadow: 0 4px 20px rgba(0,0,0,0.15);
  overflow: hidden; position: absolute;
  bottom: 70px; right: 0;
}
#oec-chat-header {
  padding: 14px 16px; background: #714B67; color: white;
  display: flex; justify-content: space-between;
  align-items: center;
}
#oec-chat-messages { flex: 1; overflow-y: auto; padding: 12px; }
.chat-msg {
  margin-bottom: 10px; padding: 8px 12px;
  border-radius: 8px; font-size: 14px; line-height: 1.4;
}
.chat-msg.user { background: #f0f0f0; }
.chat-msg.bot { background: #f3e8f0; }
#oec-chat-input {
  display: flex; padding: 8px; border-top: 1px solid #eee;
}
#oec-chat-input input {
  flex: 1; padding: 8px 12px; border: 1px solid #ddd;
  border-radius: 6px; outline: none;
}
#oec-chat-input button {
  margin-left: 8px; padding: 8px 16px;
  background: #714B67; color: white; border: none;
  border-radius: 6px; cursor: pointer;
}
</style>

services:
  # ... existing services ...

  ollama:
    image: ollama/ollama:latest
    restart: unless-stopped
    volumes:
      - ollama_data:/root/.ollama
    networks:
      - odoo-network
    # Uncomment below for GPU acceleration
    # deploy:
    #   resources:
    #     reservations:
    #       devices:
    #         - capabilities: [gpu]

docker compose up -d ollama
docker exec -it ollama ollama pull mistral  # 7B model, good balance of quality and speed

{
  "model": "mistral",
  "messages": [
    {"role": "system", "content": "Your system prompt here"},
    {"role": "user", "content": "Customer message + Odoo context"}
  ],
  "temperature": 0.3
}