
Parkey
Empowering drivers to park confidently by helping to interpret Australia's complex parking signs.

Introduction
Parkey is a start up that I worked on with a Full Stack Engineer. Since it was something we were building from the ground up, I was able to take ownership of the Product Management, Brand Design, Product Design, Front-End-Development and Digital Marketing of this product.
​
The tools I used include:
-
Figma
-
VSCode
-
Github Desktop
-
ChatGPT (for very early prototyping) / OpenAI
-
React
-
Next.js
-
Emotion CSS (for app)
-
Tailwind CSS (for website)
-
Adobe Photoshop
-
Adobe Illustrator
The problem
We identified that drivers in Sydney often find it stressful to park their cars in side streets. They are anxious about different rules at different times of days that could be misinterpreted and can result in a fine.
Our hypothesis was that Generative AI could help Australian Drivers to read complex parking rules, aiding them avoid the risk of a fine.
​
The target market we were focusing on were drivers who have come to Australia from other countries, and who aren't used to our signage. These users would struggle the most with our signs.

Product Design
I designed a simple flow along with some screens in Figma.
There were design challenges to consider, i.e., how do we know which side of the sign the user has parked? How do we inform the user that they can park, but conditionally or with a time limit?
​
We were focusing on international drivers, so I needed to make the design very clear so that drivers could clearly understand whether they were permitted to park in a given area or not, even if English isn't their first language. This is why I opted for a clear traffic light system.
​
Our goal was to get a prototype / MVP into the hands of our alpha users quickly to test.

Generative AI
The most experimental and exciting part of this project was using Generative AI. We used Open AI (GPT 4) and experimented with different prompts.
Our challenge was that it had to be accurate, otherwise users would quickly lose trust in our application.
​​
The other challenge was simply how many different rules and types of signs existed.
​
We also explored training a model ourselves with Gemini, but settled with Open AI in the end.


UI Development
Part of my role on this product included styling the components on the front end.
The project was built in React, and I used Emotion to style the components.

Branding
I chose a bright and distinctive colour palette for Parkey.
I wanted it to stand out and was aiming to create a foundation for clear, bold fonts that are accessible and easy to read.
​
The logo was designed as a 3D 'P', whose perspective made it look like a parking sign.
Website and Marketing
One of our digital marketing tactics was to start a blog with relevant topics and keywords.
I was looking for something that would help me with SEO, so I decided to use Next.js as a foundation for our website, as well as Tailwind CSS to easily style the page responsively.
​
While the Full Stack Engineer who I was working with set up the app, I was responsible for setting up and building the blog and landing page.



