https://hub.vehiclemind.com/img/vm%205.png

Vehicle Mind

Vehicle Mind is a company which sells an OBD-II device that plugs into your car and offers some useful features and insights. This includes tracking their their fuel level, battery, mileage and getting human-readable descriptions about their engine diagnostic codes when the check engine light is visible.

I joined Vehicle Mind when it was a young and small company so it had the usual traits like short deadlines, high-impact projects and each employee had a large amount of responsibility. Some of the features I worked on include:

Core Backend

Our core backend consisted of two Spring applications using DynamoDB as the database. One was serving a REST API that was being used by our mobile application and the other was ingesting the data from the devices using message queues and sending out push notifications to the user.

Device Data Sandbox

One of the tools I created at Vehicle Mind that was particularly memorable was a small web interface that could be used to view data from the test devices that we were using while developing the product. There wasn’t much complexity to the tool. It essentially let the user input the ID of the device, choose the device properties of interest, specify a date range, then view the results in a simple line graph visualization. Despite it being so basic, it ended up being very useful for debugging our test drives and lived on for far longer than expected.

Infra As Code

I had the pleasure of codifying our backend infrastructure using Terraform. This task got pretty complicated due to the sheer number of AWS services we were using and because I was aiming for a high degree of automation. Doing this made it easy for a new hire to take a fresh AWS account and get a complete copy of our backend up and running in a matter of minutes. In addition, the application was moved into docker to make it platform-independent.