Backend · 2024 · Live
WhatsApp Automation API
A backend API for handling WhatsApp auto-replies, customer conversation routing, webhook processing, and rule-based message workflows.
Client
Confidential
Proof
Live backend system with confidential conversation data removed
Timeline
3-week API and dashboard build
Industry
Backend
Year
2024
Stack
Node.js · TypeScript · PostgreSQL · WhatsApp Business API · Webhooks · Railway
The problem
Answering the same WhatsApp messages on repeat, all day, every day.
A service business was managing customer WhatsApp conversations entirely manually. The same questions about pricing, availability, and process arrived dozens of times per week. Every answer had to be typed by hand.
There was no record of past conversations, no way to track inquiry volume, and no handoff process for complex requests that needed human attention.
What we built
A backend API for WhatsApp automation that handles the repetitive work.
We built a backend API connected to the WhatsApp Business API that receives incoming messages via webhook, classifies them, and routes them to either an automated response or a human handoff queue.
- — Webhook receiver for inbound WhatsApp messages
- — Rule-based auto-reply engine
- — Message classification (auto-handle vs escalate)
- — Conversation history in PostgreSQL
- — Admin dashboard showing message volume and escalation queue
- — Configurable reply templates
How it works
Message arrives. API classifies it. Reply goes out in seconds.
When a message arrives the webhook fires and the API classifies it. If it matches a known pattern it sends an automated reply. If not, it routes to the escalation queue with a notification to staff. The entire process takes under 2 seconds.
Results
What changed.
73
messages handled automatically (%)
2
second average reply time
8
hours saved per week
Stack
Start with the audit
Want to know why customers are not contacting you?
Send your website or Instagram. We will show the first things to fix before you spend money.