WhatsApp-Based Workflow Automation Platform
Converted unstructured WhatsApp messaging into a fully structured workflow system, state-based routing, AI-powered localization, geolocation processing, and Airtable as the central data layer.
The client needed a way for users to interact with a structured workflow system without leaving WhatsApp. Incoming messages were unstructured, multi-step conversations needed to maintain state, and the platform had to support multiple languages and location-based inputs, all without a traditional app or web interface.
Client feedback“I had a very positive experience working with Silver Axis. They successfully completed the assigned project with great professionalism and patience. Throughout the process, they were responsive, attentive, and easy to communicate with. Their dedication and professional attitude made the collaboration smooth and enjoyable. I will definitely work with them again in the future.”
The transformation
What the client was dealing with, and what we replaced it with.
- All coordination happened in WhatsApp group chats, requests, assignments, and follow-ups were buried in conversation threads
- Staff manually read every incoming message and assigned it to the right person or team by hand
- No conversation state, users had to repeat context every time they sent a new message
- Volunteer matching was done manually by scrolling through availability lists and guessing availability
- Multi-language users were either ignored or required a bilingual staff member to relay messages
- Location data was shared as raw pins with no way to process or act on it programmatically
- Missed messages, duplicate assignments, and dropped requests were a regular occurrence
- Zero audit trail, no record of who was assigned what, when, or what the outcome was
- Replaced group chat chaos with a WhatsApp chatbot that handles every inbound message automatically, no human needed for routing
- State-based conversation engine remembers where each user is in a workflow, so context is never lost between messages
- Automated routing directs each request to the correct workflow path based on message content and user state
- Dynamic volunteer matching engine queries Airtable in real time to find and assign the best-fit volunteer based on availability and criteria
- Edge case handling built into every workflow, unrecognised inputs, incomplete responses, and timeouts are caught and resolved gracefully
- AI translation layer detects language and localises all outbound messages, no bilingual staff required
- Geolocation processing converts raw location pins into structured geographic data used directly in matching and routing logic
- Full audit trail in Airtable, every interaction, assignment, and outcome logged with timestamps for reporting and accountability
How we built it
- 1
Mapped all user interaction flows and designed the state machine architecture
- 2
Set up self-hosted n8n server and configured WhatsApp Business API webhook integration
- 3
Built the state-based message routing engine to evaluate and direct incoming messages
- 4
Designed Airtable schema for user data, workflow state, and request tracking
- 5
Integrated AI translation service for real-time multilingual message localisation
- 6
Implemented geolocation processing module to parse and structure location inputs
- 7
Built modular workflow blocks for each interaction sequence (registration, routing, notifications)
- 8
Tested end-to-end flows across multiple languages, location inputs, and conversation states
n8n
WhatsApp Business API
Airtable
OpenAI API
Hunter.io
HubSpot CRM
Google Sheets
Google Gemini
ClickUp
GoHighLevel
Slack