top of page

JobAssist

Personal Project

Overview

To create a AI-powered platform that can help job seekers find their dream job.
Duration: 2024. 4 - 2024. 6 (Mini case study so the duration is short) 
My role: I was the visual and interaction designer for this project.
Technologies: User research/interview, Figma, FigJam
Impact
Finished main components and screens. Plan to iterate on some details before jumping into the technical phase.

Background

Given the tough job market and frequent layoffs, job seekers are facing significant challenges in finding employment or transitioning careers. Considering the widespread popularity of AI across various industries, I aim to delve into this issue and investigate how AI can help job hunters with securing positions that align with their skills and preferences.

User Research

The user research consists of two parts: user interviews and competitor analysis. Overall, I interviewed 5 potential users (current job seekers) and then drafted the affinity diagram based on the results. I also conducted the competitor analysis, trying to spot the potential market and existing approaches. 

Interview Questions: 
1. From your point of view, why is job hunting so hard right now?
2. Which AI tools have you used during job hunting?
3. Which is your favorite AI tool, and why?
4. What are some of the improvement of current AI job hunting tools?
5. What else do you think AI can help you to achieve?

user research.png

Customer interviews result

competiter analysis.png

Competitor analysis

Customer Persona

Primary users:
End customers with age range 18 - 80, located in the United States who are currently job hunting.

User needs:
1. Be the first group of people to learn about opening roles that fit their skillsets and background.
2. More accurate AI job matching score result.
3. Interview preparation.
4. Build confident and celebrate small achievement.

Considerations

Assumptions:

1. Users have access to the internet and know how to use smart phones.

2. Users have ready-to-use resumes to be used for job applications.

3. This product signed the contracts with a group of companies, so that users can apply directly through this platform without going to companies' individual career sites. 

 

Design decisions:

1. Realistically, this product should has both desktop and mobile version to fit various needs. Focus on the mobile version for now with the assumption that people react quicker to mobile messages than desktop messages.

2. Why this app is unique than other apps should be very clear.

3. Focus on designing the features instead of the login flow.

4. Start with design related jobs, and gradually expand the scope. 

persona 2.png
persona (1) 1.png

The goal is to change a sad user to a happy user!

Sketches

After I understood the main customers painpoints and deeply analyze existing AI tool, the next step I did was to freely sketch some potential ideas to help me frame the best solutions.

WechatIMG1812.jpg

My paper sketch

Colors and Fonts

Colors and fonts are always one of the keys to the app design, because it conveys the overall mood of the app. After researching, I pick blue as the theme color. Especially for stressful job seekers, blue can calm them down so that they can focus on things that they can control. I pick SF Pro Text as the main font since this is the standard font for IOS devices. 

Main color

Main Text Color

Secondary Text Color

Notification Color

App Icon

AI Exploration

Based on my previous project experience, I conducted extensive research on designing AI experiences. Specifically, I aimed to explore AI's strengths and limitations in product design. Leveraging these research findings, I applied them to the job hunting domain.

AI.png

AI audit result

App Design

In order to solve user needs #1, what's most unique about this app is the ability to alarm users when interested companies post new roles and the convenience of bulk apply directly through the app. In the meantime, users wish to find jobs that are more likely to get callbacks from HR, so the recommendation engine needs to be highly accurate (user needs #2). This app also celebrates every small success that users achieved, helping them to gain confidence during the tough job market (user needs #4). 

The interview preparation feature is currently in the backlog. 

Featured Screens

Profile page.png

On the profile page, users can edit their job preferences and upload resumes. The AI engine uses uses' input to filter jobs and recommend matching jobs to users.

main page.png

JobAssist analyze users' preference and previous hiring data and then offers the option to apply all top picked job in one click. It saves users a lot of time manually submitting separate applications.

alert.png

Users can confirm the action before submitting all resumes to corresponding companies at once.

job page.png

With the assistant of AI and previous hiring data, JobAssist provides users with a prediction of HR callback time frame if they are successful candidates. The estimation will continue to be more accurate with more user data.

search.png

Users can search for jobs and companies using the search feature. The search engine automatically predict the input and list companies and open roles . Users can also set up job alerts with specific keywords.

App notification.png

JobAssist will send alarms to users as soon as a new fit role is posted, so users can be the first few people to apply and interview with hiring managers.

Iterations

Throughout the design process, I frequently conduct user testings using prototypes in differently fidelities. Below are some of the iterations I made based on user feedback.

1. Edit Profile Page

Initially, I designed the profile edit screen so that users could make all changes in one place without needing to navigate to another screen. Later on, I realized that there was too much happening on a single page, so I splitted potential content updates that required typeahead into a separate page. However, during user interviews, the users noted that this screen was confusing. They pointed out that all text areas appeared clickable, but only the name and title could be clicked to change the content. In order to make the design consistent, after some consideration, I made the text areas for "interested companies," "interested roles," and "interested location" also clickable, leading to individual screens for updating content.

Profile page initial.png
Profile page.png
Profile page edit.png

2. Pop-up Design Update

According to user feedback, when they saw the pop-up, they were unsure about what was being submitted to companies (resume or other materials). Therefore, as an iteration attempt, in the pop-up, I added more description explaining that their resumes would be submitted to several companies, so that users are fully aware of the action and can confirm with confidence. 

Once I show the new design to other users, I received the reaction that the system notification pop-up makes them feel that they just did something wrong and it is not personalized. Therefore as the second iteration attempt, I converted the system notification to a more personalized pop-up and also include icons and buttons on it. Further on, I iterated on the color and content of the pop-up.    

alert.png
alert edit.png
job page.png
alert.png
alert.png

Prototype (Figma Link)

Profile update + The "apply all" feature

Search for companies and jobs

Quick apply + Applied jobs

Technology

In order to successfully launch this tool, some technical blockers need to be cleared first. Below are some of the technical resources that will be helpful during the development process. 

OpenAI: A powerful tool to generate AI powered content. The development team needs to particularly focus on the prompt design, so that the job recommendation can be more accurate. 
React Native:  Open framework to develop both IOS and Android app. 
InfoTrie Job API (Or Similar): Job posting APIs to get the real-time job data from various companies. The app then notify users about new jobs from their interested companies. 

bottom of page