Refactoring and Tech Update for Parking Automation Service ZERO5

We had to research the components of the service, find weak links and fix them for ZERO5 — the parking automation service that simplifies the process of finding the garage and the parking spot.

ZERO5 offers customers access to a colossal parking network that connects drivers and garage owners, provides them with info about currently available spots, makes it easy to manage their profiles, and updates all this data in real-time.

You can find more about them here: https://www.zero5.co/

What the task was

The technologies used to develop their service have been hanging on the verge of being outdated for some time. They wanted us to refactor what’s already there and keep developing the app using modern technologies and architecture solutions.

How we solved it

Architecture

We’ve replaced SAM AWS with Serverless to make their architecture more scalable, automated, and easier to maintain and develop new features. As for the deployment process, we’ve handled AWS CI/CD to automatically create stacks with resources in AWS Cloudformation from the serverless templates.

Backend

We’ve used AWS Chime to implement in-app video calls so that drives could reach the parking service in a tap and vice versa. And AWS Gateway Websockets API to make that real-time data transfer even faster.
Moreover, we’ve added a layer of Cognito custom authorization to increase the security through all of the connected apps.

Frontend

The first thing we did after studying the project was moving it from Javascript to Typescript. It helped us dramatically reduce the number of errors, increase the overall quality of code (making it more structured, clean, and concise).

Database

We’ve researched the possibilities of migrating their database from Dynamo to AWS RDS. It would help scale the data managing process, structurize it and cut costs, as we’ve found that relational DB is the best solution for this type of product.

How it benefited our client

Refactoring has helped us shorten the project start time from several minutes to less than 20 seconds and significantly reduce errors.

Moreover, the improvements we’ve made to the product’s architecture allowed us to:

1. Increase scalability and maintenance;
2. Automate the process of deployment;
3. Increase security and protection of the data.

Our works

DemandHelm

Web platform, Media