spend-analyzer

A simple tool for analyzing spending by categorizing transaction data.

Deployment quickstart

Prerequisites

Required:

• An AWS Account
• Valid AWS credential configured and a profile set in your current shell, for example: export AWS_PROFILE=<my-profile>
• AWS API Gateway CloudWatch log role configured in the AWS Account
• An AWS Rout53 Hosted Zone
• AWS SAM cli installed
• Docker installed and running
• uv installed
• The project cloned locally

Optional:

• Request Increase AWS Account Lambda Function memory allocation increase to max 10240MB Not required but this will reduce inference times by ~20 seconds for 200 row and ~50 seconds for 500 row datasets.

---

• Create a venv

alias uv-venv='uv venv; source .venv/bin/activate'; uv-venv

First deployment requires writing the samconfig.toml and saving your configuration. Once you have a samconfig.toml you can utilize the spend-analyzer makefile.

• Build the app:

cd spend-analyzer && sam build --use-container

• Go through a guided dry-run to configure an initial samconfig.toml:

sam deploy --save-params --guided --no-execute-changeset

Complete the prompts and set values for the HostedZoneId, DomainName, and ApiDomainName template parameters.

Once the CloudFormation ChangeSet is competed you can remove the guided and no_execute_changeset keys from the [.deploy.parameters] section of your samconfig.toml

• Use the make targets to deploy the app instance:

make deploy