Back to projects

Online CV builder

Buat CV Online

Collaborative Project · QA

Case study last updated:

Project overview

Buat CV Online helps people create a CV without formatting it manually in Word. Users enter their professional background in a form, and AI generates the CV with improved wording. They can preview the result and export it as PDF or Word.

Users can also paste a job description from a job portal. AI then optimizes the CV for that target role using the supplied job requirements.

This was a two-person project with a friend who handled development. My role was QA Engineer, using exploratory testing to check the application and suggest improvements. I communicated findings through chat with screenshots or discussed them directly with the developer.

The findings below are selected examples from that work, not a complete list of everything tested or reported. They summarize observations and follow-up from exploratory testing; I did not create formal test cases or bug report documents for this project.

Role & contribution: QA Engineer, two-person collaborative project

My contribution

  • Explored the application as a user, checking how screens, forms, and navigation worked together throughout the CV creation flow.
  • Reproduced issues and explained them to the developer through chat with screenshots or in person.
  • Suggested improvements to guide users through profile sections, support additional professional links, and make page navigation clearer.
  • Retested the selected fixes and improvements after the developer applied them.

Testing scope

  • Login and registration flows.
  • Personal information, work experience, education, skills, projects, certifications, and profile-related history.
  • Job description input and ATS analysis flows.
  • Accuracy of AI-generated CV content against the user's information and target job description.
  • CV preview and PDF/Word export using the default template.
  • Mobile layouts, section navigation, pricing table readability, and running text behavior.
  • Additional profile links and URL routes when moving between pages.

Testing approach

  • Used exploratory testing to learn the application while trying different inputs, actions, and navigation paths.
  • Checked the CV creation flow from profile entry through job description input, analysis, preview, and export.
  • Observed whether controls and layouts helped users understand what to do next, including on mobile.
  • Shared reproducible findings through chat, screenshots, and direct discussions with the developer.
  • Revisited the affected flows after changes and confirmed the selected issues were resolved and suggestions were implemented.

Challenges & QA decisions

Guide users through the complete profile

I considered whether users could understand the full input flow before starting ATS optimization. Clear progression between sections mattered as much as the individual form controls.

Support different professional portfolios

I considered users outside software development, such as designers who might use Behance or Upwork. This informed my suggestion to support additional professional links.

Check navigation beyond visible content

I checked whether the URL reflected the page being viewed, as well as whether the visible page changed. This exposed a navigation issue that a visual check alone could miss.

Selected findings

Two different login dialogs served the same purpose

Observation: The application had two different login dialogs for the same function, creating an inconsistent sign-in experience.

Change made: The developer addressed the duplicate login dialog behavior.

Retest: I checked the updated login flow and confirmed the issue was resolved.

Pricing table headings overlapped the next row

Observation: Column headings extended into the row below, making the pricing table difficult to read.

Change made: The pricing table layout was corrected to keep the headings and rows separate.

Retest: I revisited the table and confirmed the overlap was resolved.

Running text did not loop continuously

Observation: The running text did not continue in an infinite loop as intended.

Change made: The running text behavior was updated to repeat continuously.

Retest: I checked the updated behavior and confirmed that it continued looping.

Profile sections lacked a clear next step

Observation: Sections such as Personal Info had no Next button. Users could start ATS optimization without noticing the other sections available to complete.

Change made: Next buttons were added to guide users between profile sections.

Retest: I checked the section flow and confirmed the new navigation was available.

Menus and sections overlapped content on smaller screens

Observation: Some menus or sections did not adapt to the available screen width and covered text below them.

Change made: The affected responsive layouts were adjusted.

Retest: I checked the affected layouts again and confirmed the reported overlap was resolved.

Professional links were limited to three fixed options

Observation: Users could only enter a website, GitHub, and LinkedIn link, with no option for other professional platforms.

Change made: Support for additional links was added following my suggestion.

Retest: I checked the additional-link feature and confirmed it was available.

Page changes did not have distinct URL routes

Observation: Navigation between pages kept the root URL instead of showing a page-specific path such as /generator.

Change made: Distinct page routes were introduced.

Retest: I navigated between pages and confirmed the URL reflected the selected page.

Results & limitations

Verified outcomes

  • All selected issues and improvement suggestions described here were addressed by the developer and retested by me.

Limitations & remaining risks

  • Template testing was limited to the default template; alternative templates were not tested.

Lessons & next steps

  • Explore how users move between sections as well as whether individual controls work.
  • Explain the observed behavior and user impact clearly, using screenshots when helpful.
  • Retest both fixes and suggested improvements after they are implemented.