From Manual Therapist Matching to an AI-Powered WhatsApp Automation System

Whenever a patient submitted the website form, their information was added to an Excel or Google Sheet.

The client then had to manually:

• Review the patient’s information
• Copy the patient’s data into ChatGPT
• Upload or provide the profiles of more than 200 therapists
• Ask AI to identify the most suitable therapist
• Review and verify the AI-generated recommendation
• Write a customized WhatsApp message
• Send the recommendation to the patient
• Update the spreadsheet to confirm that the patient had been contacted

This process required significant time and attention for every new patient.

As the number of form submissions increased, the workflow became difficult to manage and created a risk of delays, missed messages, and human error.

The client wanted a system that could automatically:

• Capture new patient form submissions
• Analyze each patient’s needs
• Compare the patient with the available therapist database
• Generate a personalized recommendation message
• Send the message through WhatsApp
• Ensure the correct recommendation was delivered to the correct patient
• Give the client visibility into incoming, outgoing, and AI-generated WhatsApp conversations

I created an automated workflow using n8n.

Whenever a patient submitted the website form, their information was added to Google Sheets. A new-row trigger then automatically started the n8n workflow.

The workflow collected the patient’s information, separated the relevant data fields, and sent the structured information to the AI system.

The AI analyzed the patient’s needs and compared them with the available therapist profiles. It then prepared a personalized recommendation message for that specific patient.

This removed the need for the client to manually copy information, upload therapist data, write messages, and track each request separately.

One of the biggest challenges was integrating WhatsApp without disrupting the client’s communication process.

When using the WhatsApp Cloud API, a business cannot always send unrestricted messages to a user outside WhatsApp’s approved communication rules. The client also needed a reliable way to continue conversations without depending on a costly third-party WhatsApp inbox.

To solve the outbound messaging limitation, I created an approved WhatsApp template message.

The first message thanked the patient for submitting the form and included an action button asking whether they wanted to proceed with receiving the information.

Once the patient selected “Yes,” the system recognized the response and automatically sent the personalized AI-generated therapist recommendation within seconds.

Another important technical challenge was handling multiple patients at the same time.

For example, ten patients could submit the form within a short period. The AI could generate ten different personalized recommendations, but each response needed to remain connected to the correct patient.

The system stored and tracked each patient’s generated response separately.

When a patient clicked “Yes” in WhatsApp, the workflow identified that specific person, retrieved their saved recommendation, and delivered only the message created for them.

This ensured that no patient received another person’s information or therapist recommendation.

The client also needed to view and manage WhatsApp conversations.

Using a third-party WhatsApp inbox would have required an additional subscription of approximately $72 to $99 per month from the beginning.

Instead, I developed a custom WhatsApp management application for the client.

Through this application, the client could:

• View incoming WhatsApp messages
• View outgoing messages
• Monitor AI-generated replies
• Send manual WhatsApp messages
• Receive customer responses
• Track conversations from one interface

I provided this custom application as an additional solution for the client, helping them avoid unnecessary third-party subscription costs.

The final workflow operated as follows:

1. A patient completed the form on the website.
2. The information was added to Google Sheets.
3. The n8n workflow was triggered automatically.
4. The patient’s information was structured and sent to the AI.
5. The AI analyzed the patient’s needs against the therapist database.
6. A personalized therapist recommendation message was generated.
7. An approved WhatsApp template was sent to the patient.
8. The patient clicked the “Yes” action button.
9. The system identified the correct patient and retrieved their saved message.
10. The personalized AI recommendation was delivered within seconds.
11. The client could monitor the complete conversation through the custom WhatsApp application.

• n8n
• Artificial Intelligence
• WhatsApp Cloud API
• Google Sheets
• Approved WhatsApp message templates
• Custom web application
• Webhooks and automated triggers
• Patient-to-therapist matching logic

The solution transformed a highly manual therapist-matching process into an automated system.

The client no longer needed to manually copy patient data, upload therapist information, generate individual recommendations, write WhatsApp messages, or update every request separately.

The new system helped the client:

• Reduce repetitive administrative work
• Respond to patients faster
• Handle multiple patient requests simultaneously
• Generate personalized messages automatically
• Reduce the risk of sending the wrong recommendation
• Manage WhatsApp conversations from a custom interface
• Avoid additional third-party inbox subscription costs
• Build a more scalable patient-to-therapist matching process

The client was extremely satisfied with the completed automation because it simplified the entire workflow while maintaining personalized communication with every patient.

This project demonstrates how AI automation can improve more than just messaging.

By connecting the website form, therapist database, AI matching process, WhatsApp communication, and conversation management system, I created a complete operational workflow for the client.

What previously required several manual steps could now be completed automatically within seconds, allowing the client to focus on supporting patients rather than managing repetitive administrative tasks.