CVMakeover.AI

Inspiration

As graduate students actively applying to internships and full-time roles, we spend a surprising amount of time rewriting resumes. Every job description highlights a slightly different mix of skills, tools, and expectations, which forces applicants to repeatedly tweak bullet points, reorder sections, and rephrase experience - even though the underlying work has not changed.

This repetitive process quickly becomes inefficient and mentally exhausting, especially when applying to many roles in parallel. We were inspired to build CVMakeover.AI as a way to reduce this friction and make resume customization faster, more consistent, and more intentional.

Our core belief is simple:
Your experience stays the same - only the way it is framed should adapt to the job.

What it does

CVMakeover.AI helps users generate tailored, ATS-friendly resumes based on a single structured profile and a target job description.

Instead of rewriting resumes from scratch, users:

Maintain one structured profile containing their real education, experience, projects, and skills
Provide a job description as input
Receive a customized resume that emphasizes relevant skills and keywords without fabricating experience

The final output is a clean LaTeX-based resume, ready to be compiled into a professional PDF.

How we built it

We designed CVMakeover.AI as a modular system with clear separation between frontend, backend, and presentation layers:

Chrome Extension (Frontend)
Allows users to interact directly with job postings while browsing and trigger resume generation from the page.
FastAPI Backend
Handles job description processing, profile tailoring, and resume generation.
Structured Profile (JSON)
Acts as a single source of truth for a user’s experience, making customization reproducible and consistent.
Optional LLM Integration
When enabled, an LLM rewrites resume bullets to better match job-relevant language while remaining strictly truthful.
LaTeX + Jinja2 Templates
Ensures professional formatting, consistency, and ATS-friendly output.

This architecture allows the frontend and backend to be developed independently while still working together as a single system.

Challenges we ran into

One of the biggest challenges was balancing automation with authenticity. Resume optimization should improve clarity and relevance without exaggeration, so we had to carefully design prompts and safeguards around AI-assisted rewriting.

We also faced challenges in:

Managing multiple development environments (browser extension + backend)
Handling unstructured and subjective resume data
Coordinating work across branches while developing in parallel
Ensuring the system remained useful even before all features were complete

These challenges pushed us to be more disciplined about structure, testing, and incremental progress.

Accomplishments that we're proud of

Building an end-to-end system that connects a browser extension to a backend service
Designing a reusable resume profile format that scales across roles
Generating production-quality LaTeX resumes programmatically
Integrating optional AI assistance while preserving truthfulness
Creating a tool we genuinely want to use ourselves

What we learned

Through this project, we learned how to:

Design AI-assisted systems with ethical constraints
Structure real-world user data for automation
Build clean APIs that support evolving frontends
Debug complex development workflows involving multiple tools and environments

Most importantly, we learned that even small workflow improvements can have a large impact when applied to problems people face repeatedly.

What's next for CVMakeover.AI

Going forward, we plan to:

Improve job description parsing and keyword relevance
Add resume feedback and match scoring
Support multiple resume templates and formats
Enhance the extension UI for smoother user experience
Prepare the project for deployment beyond local development

CVMakeover.AI started as a personal productivity idea, but we see strong potential for it to become a practical tool for job seekers who want to apply smarter - not harder.