RUNNING_TERRAFORM_AWS.md

Running with Terraform and AWS

Creating AWS resources with Terraform

1. First of all, you need to have a AWS account configured (with the all necessary permissions) locally:

aws configure

2. Go to infrastructure directory

cd infrastructure

3. Create a Terraform backend config file

cat > aws-backend.conf <<EOF
bucket = "your-tfstate-bucket"
key    = "state/aws.tfstate"
region = "sa-east-1"
EOF

Change the bucket, key and region as you want.

4. Initialize the Terraform

terraform init -backend-config=aws-backend.conf

5. Create all resources with terraform apply

terraform apply -auto-approve

If you want, execute a terraform plan before.

Configuring timeless-api (backend application)

1. Go to timeless-api directory

cd timeless-api

2. Create a .env file

cat > .env <<EOF
AWS_ACCESS_KEY_ID=
AWS_SECRET_ACCESS_KEY=
OPENAI_API_KEY=
SECURITY_KEY=
INCOMING_MESSAGE_FIFO_URL=
RECOGNIZED_MESSAGES_FIFO_URL=
EOF

3. Fill in all environment variables

The table below explains the purpose of each variable along with suggested example values:

Variable	Description	Suggested Example
AWS_ACCESS_KEY_ID	Your AWS Access Key	AKIAxxxxxxxxxxxxxxx
AWS_SECRET_ACCESS_KEY	Your AWS Secret Access Key	xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
ASSETS_BUCKET	The name of the bucket where audio and images will be stored (created via Terraform)	my-assets-bucket
OPENAI_API_KEY	Your OpenAI API key	sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
SECURITY_KEY	Secret key used for AES encryption (must be exactly 16, 24, or 32 characters long, no accents)	MySuperSecretAESKey12345678901234
INCOMING_MESSAGE_FIFO_URL	The URL of the SQS FIFO queue for incoming messages	https://sqs.us-east-1.amazonaws.com/123456789/incoming-messages.fifo
RECOGNIZED_MESSAGES_FIFO_URL	The URL of the SQS FIFO queue for processed messages	https://sqs.us-east-1.amazonaws.com/123456789/messages-processed.fifo

4. Execute the application in Dev mode

./mvnw quarkus:dev -Dquarkus.profile=aws

5. Sign up at http://localhost:8080/sign-up

6. Sign in at http://localhost:8080

7. Add the user phone number at http://localhost:8080/home/user-configs, this phone number must be set after on ALLOWED_PHONE_NUMBERS.

Configuring and running the whatsapp application

1. Go to whatsapp directory

2. Install packages

npm install

3. Set the following environment variables in your .env file:

Variable	Description	Suggested Example
AWS_ACCESS_KEY_ID	Your AWS Access Key	AKIAxxxxxxxxxxxxxxxxxx
AWS_SECRET_ACCESS_KEY	Your AWS Secret Key	xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
ASSETS_BUCKET	The name of the bucket where audio and images will be stored (created via Terraform)	my-assets-bucket
ALLOWED_PHONE_NUMBERS	Comma-separated list of phone numbers allowed to interact with the bot	5511999999999,5511888888888
OPENAI_API_KEY	Your OpenAI API Key used to access GPT and Whisper APIs	sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
INCOMING_MESSAGE_QUEUE	URL of the SQS queue that receives incoming messages	https://sqs.us-east-1.amazonaws.com/123456789/incoming-messages.fifo
RECOGNIZED_MESSAGES_FIFO_URL	URL of the SQS FIFO queue where processed messages are sent	https://sqs.us-east-1.amazonaws.com/123456789/messages-processed.fifo

4. Start the application

npm run start

3. Open the WhatsApp and scan the QR Code generated

After configuring and running all applications

How to Use the Application

1. Open the sign-up page at http://localhost:8080/sign-up.

2. Fill out the registration form with your details and submit it.

3. Log in to your account at http://localhost:8080.

4. Add your phone number on the User Configs page.
   ⚠️ Make sure the phone number matches one of the values defined in the ALLOWED_PHONE_NUMBERS environment variable.

5. After connected the device with the scanned QR Code, send the following message: "Achei mil reais no chão da praia".

   You should receive something like it:

   

6. Refresh the page and see your balance go negative — good luck managing your expenses!