Background

Gojara initially launched as an e-learning platform aimed at job-seekers, with job centers being the primary clientele.

However, shortly after its launch, the founders opted to expand access directly to end-users. Along with this strategic pivot, they introduced significant new features, transitioning the platform from a singular focus on e-learning to a versatile 3-in-1 resource for job seekers.

Time Frame

Speed to market was a priority for my client, who aimed to launch a Minimum Valuable Product (MVP) swiftly to showcase to potential investors.

The project was allocated a total of 38 hours, requiring efficient use of time to deliver a functional and presentable MVP within this tight schedule.

=
1
E-learning for Job Seekers
  • Find out what you really want from your next job.
  • Learn how to craft a perfect application.
  • Learn how to ace the job interview.
  • How to find the right job offerings.
Initial Platform
2
CV Maker
  • An intuitive tool designed to assist you in creating an outstanding CV.
  • Select a professional layout for optimal presentation.
  • Manage multiple CVs.
Built by another team
3
Application Builder
  • Create individual Applications with just 4 clicks.
  • Pick your application strategy.
  • Supported by AI.
Main focus of this story

Project Brief

Having previously collaborated with the client on restructuring their e-learning content, I was approached to develop a concept for the Application Builder.

This tool is designed to assist job seekers in swiftly crafting targeted applications for specific job openings, leveraging the capabilities of GPT technology.

My Role

As a UX/UI Consultant, my responsibilities included designing user flows, wireframes, and high-fidelity screens, along with creating a straightforward click-through prototype for its demonstration.

Additionally, I focused on creating and refining the prompts for GPT to generate personalized, high-quality cover letters.

Results

Possible Investors

We successfully launched the MVP on schedule, featuring both the Application Builder and the CV Maker. This timely launch enabled the client to confidently engage in multiple significant investor pitching rounds. Although still in progress, the initial reception from these rounds has been very promising.

Gathering Data

While the project is currently available to the public, it has not yet been actively marketed. We are in the process of gathering comprehensive data, which will be crucial for further refining and enhancing our platform as we prepare for a wider public release

Our Approach

Introduction

Due to the urgency of having a functional live version as quickly as possible, we adopted an agile approach for this project, swiftly transitioning from concept to final solutions. This approach involved making pragmatic decisions along the way.

General Solution Process

An image depicting the overall process: Phase 1 involves continuous ideation, prototyping, and testing. Phase 2 is focused on Design & Build. The final step is measuring results.An image depicting the overall process: Phase 1 involves continuous ideation, prototyping, and testing. Phase 2 is focused on Design & Build. The final step is measuring results.

Site Structure & User Flow

Sitemap Modifications

The previous sitemap was initially designed solely with the focus on E-learning.

A hub has been added in the new version, which can be accessed via the main navigation. The hub serves as an overview and administration page for the two new features, the CV Maker and the Application Helper, while the main e-learning page remains Home.

A diagram showing the platform sitemap. The Hub for th CV Maker and Application builder is attached to the E-Learning platform.
The platform's sitemap reveals that the Hub, including the CV Maker and Application Builder, operates as a parallel tool accessible through the Gojara E-learning platform.

Mapping the flow

Introduction

The Application Builder will lead users through a series of steps to create an application that not only aligns with a particular job posting but also showcases the candidate's skills and aspirations effectively.

To discuss this process at a broader level, I created an initial User Flow Diagram as a reference for our team.

High-Level Flow

This flow highlights that in order to use this tool, the user should have already made at least one CV using the CV maker.

To Application maker flow itself is substituted by a grey box to simply this flow.

Creating an Application - Flow

The next flow provides a more detailed view of the grey box from the previous flow. It demonstrates the specific steps the user needs to take to create a personalized application, including system checks and waiting times.

Initial High-Fidelity
Mock-ups

The team responsible for the CV Maker had already created a fundamental Style Guide.  This made it easy for me to blend both tools together.

I used their existing parts and added new ones as needed. So, I moved straight into a more detailed version.

I want to discuss three of these screens in more detail

1

The Hub

The hub serves as an overview and administration page for CVs and applications. Users can also upload other important documents, such as certificates or job references, to add them to their applications quickly and easily.

Creating a new application is the main action on the page, as this is usually the most common usage.

2

First Screen of the Application Builder

When designing the initial screen for the Application Builder, simplicity was our main goal.

In this first version, we only intended to offer the input of a URL. The user is only shown a text field for the direct insertion of a job advertisement if the analysis of the URL fails.

This step should be perceived as a kick-off that initiates the process and not as the first step in a sequence of steps, which is why we have decided not to show it as part of the progress bar.

3

Choose your Application Strategy

Building trust in the output of an AI system is essential for a positive User Experience. That's why the 'Strategy' step is key. At this stage, users select their application strategy.

The AI suggests a strategy tailored to the user, drawing on the job description, the CV's details, and insights from e-learning activities. The recommended AI strategy is laid out on the left, demonstrating to the user how it's been personalized for their specific situation.

Screen and Flow Optimization

Lightweight User Testing

The client used the prototype I created to conduct a round of usability tests. The tests revealed a number of problems, particularly with regard to wording. The most comprehension problems for users occurred during the download process of the finished application.

In Team Discussions

During team discussions, I collaborated with the senior developer to thoroughly analyze each step, aiming to identify potential risks and opportunities for quick wins and enhancements.

Some of the changes we made:

1

Flow Reorganization

We moved the user's CV selection to the start of the process to use it as the foundation for AI analysis, especially when multiple CVs are present.

However, this change removed the immediate satisfaction of witnessing the AI in action, which can be a significant drawback.

To mitigate this, we introduced an animated banner at the top of the screen to indicate that the AI is actively working in the background. Further testing is required to ensure its effectiveness.

2

Sometimes an extra step can be beneficial

Initially, I had planned for the application to be automatically downloaded to the user's download folder, with a success banner displayed upon completion, returning the user to the Hub.

However, this approach proved confusing for many testers who were unsure that something had been downloaded and where to locate it.

To address this, we introduced an additional step where the user specifies the location for the download to be saved.

3

Simplicity can have its drawbacks

Initially, the idea of having just a URL field on the Application Builder's entry page was appealing, but it ultimately presented some challenges:

  • Not everyone is familiar with what a URL is.
  • Job portals or websites with job listings vary significantly and aren't always easily parsed..
  • This approach excludes job offers that users might have received through email, messenger apps, and other means.
A High Fidelity version of the first version of teh JobUpload, with just an input field for the URL
First concept, just showing the form field to enter an URL.
A High Fidelity version of the second version of the Job Upload, with  an input field for the URL and a text post to paste the text.
First iteration, showing an additional text box to enter job descriptions manually-

Prompting for GPT

Gojara employs the latest GPT model for analyzing job descriptions, developing a well-structured application strategy, and generating cover letters.

To achieve this, I employed various prompting techniques that collaborate to produce the final result. This involves multiple iterations of persona prompting, where GPT assumes specific roles before tackling assigned tasks.

While our current version shows significant promise, we anticipate continuous refinement and improvement over time.

NDA

I'm unable to disclose any specific details regarding the prompts and the methodology we employ to generate the analysis and create the cover letter.

Live Testing & Feedback

Introduction

After the product launch, Gojara had the opportunity to carry out extensive tests with B2B customers. Due to the high response and excellent feedback, Gojara has several opportunities for cooperation.

During these tests, a major concern was consistently raised, and this concern was substantiated by the data collected by Gojara.

Reorganising the Architecture

Initially, for practicality reasons, we integrated the CV Maker and the Application Builder into the e-learning platform, with the latter serving as the central component.

However, this approach resulted in user confusion, as they expected a 3-in-1 tool but primarily encountered the e-learning aspect.

To address this issue, we made the decision to restructure the platform by introducing a Main Page that serves as a jumping-off point to the various sections of the platform. This change aimed to provide users with a clearer and more intuitive navigation experience.

Main Page Concept

The primary emphasis for the Main Page should revolve around the application builder and the e-learning experience.

We also aimed to provide users with clear instructions on how to use the application builder, specifically by emphasizing the need to create at least one CV using the CV maker.

Conclusion

Witnessing the meaningful growth of a platform is truly inspiring. The positive feedback we've received from users and potential investors is highly motivating.

The adjustments made to the site structure underscore the significance of not only testing new features but also how they integrate with the existing framework.

In light of the urgency to bring the product to market swiftly, we acknowledge that what we have now is a Minimum Viable Product (MVP). We recognize the need for ongoing improvements in the future to refine and enhance the platform.

Explore another story