WhatsApp Integration Workflow in N8N (File handling, Postgres Table, File Bucket, Sub workflows)
Budget: $50 – $100 USD
I am looking for a WhatsApp integration workflow in n8n (saas, not self hosted). The workflow will be connected to a WhatsApp business account, so that it can respond to chat messages. (Just standard nodes, no community is preferred)
It must be able to handle:
1. Attachments (PDF, TXT, CSV, JSON) => Store them converted in predefined tables + a storage bucket in original formats with sessionid+filename.fileextension (There are 2 buckets and 2 tables: JD and CV.)
2. Inputs should be either typed text, or transcribed audio. No voice processing is needed, except what is built in the keyboard transcribe feature.
3. Conversation about other tables and vectors from the Postgres DB
4. Execute a canonical step sequence, activating several sub workflows which handle specific tasks and return the results to the Orchestrator agent.
4.1. Sub workflows are:
4.1.1. Questioning: it generates questions based on certain tables. Creating a 3rd table questions and answers.
4.1.2. Reporting: It generates a report about the results of the questioning
The result should be: A main orchestrator workflow handling the Whatsapp chat inputs and attachments, holding the conversation and creating files in the bucket and updates in the tables. Plus 2 sub workflows handling the specific tasks reading all relevant tables.
Related templates are:
1. AI-Powered WhatsApp Chatbot for Text, Voice, Images, and PDF with RAG | n8n workflow template
2. AI Agent To Chat With Files In Supabase Storage and Google Drive | n8n workflow template
Related video for Supabase bucket: Supabase Storage and N8N How To! - YouTube
Success criteria:
1. I only need to import the template and add my credentials as in each standard template of n8n (if you don’t know, try importing one and you will see there is a prompt to add your credentials before the import.)
2. All functions above are working with all file types
3. A webhook version is delivered replacing WhatsApp for dedicated application use (Which is basically the same just the download and get file mechanism is different.
4. A simple how to adjust the template document is delivered for:
a. Replacing the dummy parsing
b. Adding additional tools, sub workflows and how the system prompts of each node must be adjusted to fill it with life
5. I can successfully test it in my own instance with you seeing it. Shared screen.
It must be able to handle:
1. Attachments (PDF, TXT, CSV, JSON) => Store them converted in predefined tables + a storage bucket in original formats with sessionid+filename.fileextension (There are 2 buckets and 2 tables: JD and CV.)
2. Inputs should be either typed text, or transcribed audio. No voice processing is needed, except what is built in the keyboard transcribe feature.
3. Conversation about other tables and vectors from the Postgres DB
4. Execute a canonical step sequence, activating several sub workflows which handle specific tasks and return the results to the Orchestrator agent.
4.1. Sub workflows are:
4.1.1. Questioning: it generates questions based on certain tables. Creating a 3rd table questions and answers.
4.1.2. Reporting: It generates a report about the results of the questioning
The result should be: A main orchestrator workflow handling the Whatsapp chat inputs and attachments, holding the conversation and creating files in the bucket and updates in the tables. Plus 2 sub workflows handling the specific tasks reading all relevant tables.
Related templates are:
1. AI-Powered WhatsApp Chatbot for Text, Voice, Images, and PDF with RAG | n8n workflow template
2. AI Agent To Chat With Files In Supabase Storage and Google Drive | n8n workflow template
Related video for Supabase bucket: Supabase Storage and N8N How To! - YouTube
Success criteria:
1. I only need to import the template and add my credentials as in each standard template of n8n (if you don’t know, try importing one and you will see there is a prompt to add your credentials before the import.)
2. All functions above are working with all file types
3. A webhook version is delivered replacing WhatsApp for dedicated application use (Which is basically the same just the download and get file mechanism is different.
4. A simple how to adjust the template document is delivered for:
a. Replacing the dummy parsing
b. Adding additional tools, sub workflows and how the system prompts of each node must be adjusted to fill it with life
5. I can successfully test it in my own instance with you seeing it. Shared screen.
Related categories:
Data Processing
PostgreSQL
Automation
Chatbot
Database Management
API Integration
Data Management
SaaS
AI Chatbot
n8n