Creating an AI Resume Builder with Firebase Studio

For the longest time, I've always wanted to create a resume builder, similar to Jobscan or something similar. One of my childhood buddies had created a resume builder for himself and I asked Firebase Studio.

Firebase Studio is a cloud-based, agentic development environment that simplifies building full-stack AI apps. It allows developers to quickly prototype, build, deploy, and run apps using a web-based interface and generative AI. It includes tools like emulators, testing frameworks, and an AI assistant powered by Gemini. 

To create the resume builder, I used the prompt below:

use angular to create a resume builder similar to this https://www.designbymobius.ca/resume. Use color theme: FFFBDE, 90D1CA, 129990, 096B68

Sample Resume Builder created by Firebase Studio

Here are the features created that I think were nice:

  • Fields are fillable and changes appear on the resume preview in real time.
  • The resume style can be changed from the dropdown menu with the current selection "Modern"
  • The "Save Progress" functionality works as expected.

The basic functionalities worked such as filling/updating the resume fields, however the "Print/PDF" button only displayed a toaster message when clicked.

While this was nice I wanted to also create a resume optimizer that matched for keywords in job descriptions. In addition, it will provide suggestions for keywords to add an uploaded resume. Note: I had done some research using NotebookLM on resume builder pain points from users online, and also asked prompted it to provide me with a proposed MVP recommendation which I prompted to Firebase Studio. Here's a snippet of the prompt:

A simple, buildable MVP could focus on the following:

1.Accurate Resume Import (PDF/Word): Allow users to upload their existing resume. The primary focus here is on accuracy in parsing essential sections (contact info, summary, experience, education, skills) and, most importantly, absolutely not forcing templates or deleting user content [1]. Basic editing of the parsed text should be enabled [1].

2.Simple Job Description Input: Allow users to easily copy/paste or upload a job description [1].

3.Core Keyword Matching & Highlighting: Implement the core function of scanning both the resume and the job description to identify and highlight relevant keywords. Clearly show which relevant keywords from the job description are missing from the resume [1, 2]. Keep the initial keyword matching logic straightforward.

ATS Resume Optimizer Screenshot

The application was pretty decent. It highlighted matching keywords in green and non-matching in red. It also had a calculated score to determine the percentage keyword match.

Resume Optimizer prototype for ATS

While both these applications look decent at a prototype level, I'm going to attempt merging these 2 applications into one. Here are the other improvements I want to make:

  • The resume builder saves the filled resume in some storage
  • The Resume Optimizer can fetch stored resumes and then proceed to do the ATS keyword matching.
  • The AI Keyword Suggestions tracks for generic words like "about". I want to narrow the dictionary to words specific to the job domain.
  • Fix the "Print PDF" functionality

I'm really impressed with what Firebase Studio was able to implement right off the bat. It provides a good starting point for me to code and prompt additional improvements.