How to Send and Receive SMS with Make.com Using Your Android Phone

You can build an AI agent in Make.com that receives customer text messages, understands what the sender is asking, and automatically sends an SMS reply — while continuing to use the Android phone number and SIM you already have.

In this setup, the Android phone remains the SMS channel. AutoForward Text connects the phone to Make.com, while a Make AI Agent processes the conversation and decides what the response should say.

This gives you a practical way to build an SMS AI agent for customer enquiries, appointment questions, lead qualification, after-hours responses, support requests, and other business conversations without moving the phone number to a separate cloud SMS provider.

How the SMS AI Agent Works

There are four main parts to the workflow:

1. Your Android phone receives an SMS.
2. AutoForward Text sends the message to a Make webhook.
3. Make AI Agent reads the message and generates a response.
4. Make posts the response back to AutoForward Text, which sends the reply through the Android phone.

The AI does not send the SMS itself.

Think of each component like this:

Component	Responsibility
Android phone	Sends and receives SMS using the SIM and mobile number.
AutoForward Text	Connects the Android SMS channel to webhooks and reply APIs.
Make.com	Runs the automation scenario.
Make AI Agent	Understands the customer’s message and decides what to say or do next.

Before You Start

You will need:

• An Android phone with an active SIM
• SMS service on that SIM
• AutoForward Text installed and connected on the Android phone
• SMS permission enabled
• A Make.com account

Step 1: Connect Your Android Phone

Install the AutoForward Text Android app and sign in with your AutoForward Text account.

Grant SMS permission when prompted. Then disable battery optimization for AutoForward Text so Android can keep the app running reliably in the background.

On phones with aggressive battery management, also check sleeping apps, background restrictions, auto-start settings, and manufacturer-specific power-saving options.

Open the AutoForward Text dashboard and confirm that the Android device appears as connected.

Enable SMS Gateway

Receiving messages is only half of the AI agent workflow. To allow Make to send an SMS reply back through the phone, enable SMS Gateway for the connected Android device.

Before configuring Make, send a normal SMS from another phone to the Android number:

Confirm that the incoming message appears in the AutoForward Text message log. If it does, the Android side is ready.

Step 2: Create a Custom Webhook in Make

Create a new scenario in Make.com.

Add:

Create a new webhook and give it a clear name such as:

Make generates a unique webhook URL. Copy that URL. AutoForward Text will send each incoming SMS to it.

Step 3: Forward Incoming SMS to Make

Return to AutoForward Text and open the forwarding settings for the Android device.

Add or enable a Webhook destination.

Configure it with:

• Method: POST
• Webhook URL: your Make Custom webhook URL
• Source: incoming SMS from the selected Android device

The webhook should include the useful message fields your Make scenario needs, such as:

• event
• device_id
• device_name
• contact_name
• contact_number
• content
• direction
• timestamp_utc
• message_id
• reply.endpoint

The reply.endpoint is especially useful. Make does not need to reconstruct which Android device, number, or message should receive the response. It can send the generated answer directly to the reply endpoint supplied with that incoming SMS.

A simplified payload may look like this:

{
  "event": "sms.received",
  "device_id": 36664,
  "device_name": "Office Android",
  "contact_name": "Customer",
  "contact_number": "+14155550123",
  "content": "Hi, are you available tomorrow?",
  "direction": "incoming",
  "message_id": 9274,
  "reply": {
    "method": "POST",
    "endpoint": "https://www.autoforwardtext.com/api/.../reply"
  }
}

Step 4: Capture a Real SMS in Make

Return to Make and click Run once.

Send a fresh SMS to your Android phone:

The message should arrive on the Android phone, appear in AutoForward Text, and then trigger the Make webhook.

Open the webhook output bundle and confirm that you can see contact_number, content, device_id, message_id, and reply.endpoint.

Step 5: Add Make AI Agent

Add a new module after the webhook:

Connect the AI provider or model you want to use.

For the agent input, map the SMS content from the webhook:

The AI Agent will now receive the customer’s SMS as the message it needs to understand and answer.

Step 6: Give the Agent Clear Instructions

Do not start with a complicated autonomous agent. The first version should have a narrow job and strict boundaries.

You are the first-line SMS assistant for our business.

Answer the customer's question clearly and briefly.

Do not invent prices, policies, opening hours, availability,
stock levels or other business facts that have not been
provided to you.

If you do not know the answer, say that someone from the team
will follow up.

Keep normal replies below 60 words.

SMS replies should sound natural and conversational.

Do not claim to be a human.

Then give the AI Agent the business information it actually needs, such as products, services, pricing, opening hours, areas served, FAQs, appointment rules, and escalation rules.

Step 7: Give Each SMS Conversation Memory

Conversation memory matters because customers rarely restate the entire context in every SMS.

Use the sender’s phone number as the conversation identifier:

This keeps messages from the same sender in the same AI conversation while separating different customers.

Step 8: Create an AutoForward Text API Key

Open API Keys in AutoForward Text and create a new key with a name such as:

Give the key the permission required to reply to messages.

Copy the API key when it is shown and store it securely.

Step 9: Send the AI Response Back by SMS

Add an HTTP module after the AI Agent:

Method

URL

Do not hard-code a message ID or reply URL. Each incoming SMS should use the reply endpoint provided with that message.

Authorization

Authorization: Bearer YOUR_AFT_API_KEY

Idempotency

Add an Idempotency-Key based on the incoming AutoForward Text message ID:

Idempotency-Key: make-sms-9274

In the live scenario, map message_id dynamically.

Content Type

Content-Type: application/json

Request Body

{
  "text": "AI AGENT RESPONSE"
}

Map the Make AI Agent output into text.

Step 10: Test the Complete SMS AI Agent

Send a fresh test message and verify each stage:

1. The Android phone receives the SMS.
2. AutoForward Text logs it.
3. AutoForward Text POSTs it to Make.
4. The Custom webhook triggers the scenario.
5. Make AI Agent reads the SMS.
6. The agent generates an answer.
7. The HTTP module POSTs the answer to reply.endpoint.
8. AutoForward Text routes the reply to the Android device.
9. The Android phone sends the SMS.
10. The customer receives the response.

Once the scenario works reliably, turn it on so incoming SMS webhooks are processed automatically.

Make the Agent Useful, Not Just Automatic

A useful business SMS agent needs reliable knowledge, short responses, and a clear human handoff path.

Give It Real Business Knowledge

Provide the facts it is expected to answer: services, pricing, availability rules, locations, policies, and common questions.

Keep SMS Responses Short

Instead of a long email-style answer, prefer something like:

Short replies are easier to read and feel more natural over SMS.

Know When to Escalate

Make can then create a helpdesk ticket, Slack alert, CRM task, email notification, or human approval step.

Optional: Give the Agent Tools

Once the basic SMS loop is reliable, you can give the AI Agent selected tools.

• Check an appointment in a calendar
• Look up an order status
• Create or update a CRM contact
• Create a callback task
• Open a support ticket

Practical SMS AI Agent Ideas

Use Case	Example Workflow
Lead qualification	Prospect texts a question → AI answers → collects useful details → Make creates or updates the CRM lead.
Appointment assistant	Customer asks for Friday afternoon → AI checks scheduling rules or calendar availability → replies with the next step.
After-hours receptionist	Customer texts outside office hours → AI answers basic questions → captures the reason for contact → creates a follow-up task.
Field service	Customer reports an issue → AI collects essential details → Make creates a service request or technician task.
Basic customer support	Incoming SMS → AI answers from approved business knowledge → anything uncertain is escalated to a person.

What About Make.com’s Android Integration?

Make.com also offers Android-related modules, including SMS actions, which may be enough for some simple Make-only automations.

The AutoForward Text approach is more useful when you want the Android SMS number to sit inside a broader messaging setup where messages can be logged, forwarded through webhooks, replied to through APIs, and connected to systems beyond Make.

Do You Need Twilio?

No, not for this setup. The connected Android phone and its SIM are the SMS endpoint.

A traditional cloud SMS provider like Twilio may be better when you need large message volumes, multiple virtual numbers, fully cloud-based infrastructure, or more sophisticated telecom routing.

Troubleshooting

Problem	Likely Cause	What to Check
SMS does not appear in AutoForward Text	Android app, SMS permission, or background restriction	Check app connection, SMS permission, internet, and battery settings
SMS reaches AFT but Make does not run	Webhook rule or URL problem	Confirm webhook forwarding is enabled and the Make webhook URL is correct
Make receives the SMS but AI does not respond	Input mapping or model connection	Check that content is mapped into the agent input and the AI model is connected
AI answers correctly but no SMS arrives	HTTP request, API key, gateway, or Android problem	Check HTTP response, Authorization header, SMS Gateway status, phone internet, SIM service, and SMS permission
AI forgets earlier messages	Conversation ID missing	Map contact_number into the Conversation ID
Different customers share context	One Conversation ID reused	Use a customer-specific value such as the sender’s number
Duplicate replies	Scenario retry created another send	Use one stable Idempotency-Key per incoming message
Messages stop after the phone is idle	Android battery optimization	Exclude AutoForward Text from battery saver and sleeping-app restrictions

Frequently Asked Questions

Can Make.com receive SMS from my Android phone?

Yes. AutoForward Text can forward incoming Android SMS messages to a Make Custom webhook, which then triggers the Make scenario.

Can Make AI Agent automatically answer SMS?

Yes. Map the incoming SMS into Make AI Agent, generate the answer, and send that response through the AutoForward Text Reply API for delivery by the Android phone.

Do I need a Twilio number?

No. This workflow uses the connected Android phone and its SIM as the SMS channel.

Which Make modules do I need?

The basic workflow uses a Custom webhook, Make AI Agent, and an HTTP request module.

Can the SMS AI Agent remember previous messages?

Yes. Use the sender’s phone number as the Make AI Agent Conversation ID so later messages from that number remain part of the same conversation.

Can the AI Agent access my CRM or calendar?

Yes. You can connect additional Make modules or tools so the agent can look up or act on approved business data.

Can I send a new SMS rather than replying to one?

Yes. A separate AutoForward Text Send API workflow can initiate a new SMS through a gateway-enabled Android device. The reply endpoint is intended for responding to a specific incoming message.

Does my Android phone have to remain online?

Yes. The physical phone sends the SMS, so it needs power, internet access, SMS permission, mobile service, and permission to run AutoForward Text in the background.

Is Make AI Agent required?

No. You can replace the AI Agent with filters, fixed responses, CRM lookups, routers, databases, or other Make modules.

Is this suitable for bulk SMS marketing?

No. The intended use is legitimate operational and conversational messaging. Carrier rules, consent requirements, local laws, anti-spam requirements, and mobile-plan limits still apply.

Final Thoughts

The useful part of this workflow is not simply that AI can write an SMS response.

It is that an existing Android phone number can become an input and output channel for an AI automation.

AutoForward Text handles access to the SMS channel. Make handles the workflow. The AI Agent handles the conversation.

Start with a narrow agent that can answer a small set of questions safely. Make sure the receive-and-reply loop works reliably. Then add memory, knowledge, CRM data, tools, qualification logic, appointment handling, and human handoff as needed.