Brieflyy

• 

  Brieflyy's clean and intuitive home page introduces users to AI-powered CV tailoring with full human approval.

• 

  Users paste their CV, select a target role, and provide their API key to begin the tailoring process.

• 

  Brieflyy analyzes the job requirements and highlights the user's matching strengths and relevant skills.

• 

  Compare the original CV with the AI-tailored version and users review and approve each change individually before generating final CV.

• 

  Application submitted.

• 

  The approved CV is ready to copy and use for applications, ensuring transparency and accuracy throughout the process.

Inspiration

As computer science students, we've spent countless hours rewriting our CVs for every internship application. Every company emphasizes different skills and keywords, making the process repetitive, time-consuming, and stressful.

While AI tools exist, many automatically generate content and sometimes exaggerate or invent experience. We wanted something more transparent and trustworthy.

That's why we created Brieflyy: an AI-powered CV tailoring assistant that helps users adapt their CVs to a target role while keeping them in control of every change.

What it does

Brieflyy helps users customize their CVs for specific job opportunities in seconds.

Users simply:

• Paste their CV
• Select a target job
• Let AI analyze the role and identify relevant requirements
• Receive a tailored version of their CV
• Review every modification through a side-by-side diff view
• Approve all changes before generating the final CV

Unlike many AI writing tools, Brieflyy never invents new skills or experiences. It only rephrases and reorganizes existing information to better match the selected role.

How we built it

Brieflyy is a React and Tailwind web application created with Lovable.

The application uses Anthropic's Claude API to analyze job descriptions and rewrite CV content while following strict instructions that prevent fabricated information.

The workflow includes:

• CV input and job selection
• AI-powered job analysis
• Generation of matched strengths
• Line-by-line comparison between the original and tailored CV
• Mandatory user approval before finalizing the document

Sample job postings are currently provided through a static JSON dataset.

Challenges we ran into

One of the biggest challenges was preventing hallucinations. AI models naturally tend to "improve" text by introducing information that wasn't originally provided.

To address this, we designed strict prompt constraints and implemented a mandatory review system where users must approve every change before continuing.

Another challenge was presenting AI output in a transparent way rather than as a black box.

Accomplishments we're proud of

• Building a fully functional AI-powered CV tailoring workflow
• Designing a human-in-the-loop approval system
• Creating a side-by-side diff interface for complete transparency
• Ensuring the AI never fabricates skills or experiences

What we learned

We learned that trust and transparency are just as important as AI capabilities.

Building responsible AI products requires more than generating text—it requires keeping users informed and in control.

We also learned how careful prompt engineering can significantly improve reliability and user confidence.

What's next

• Support PDF and DOCX uploads
• Integrate real job board APIs
• Allow users to save multiple CV versions
• Provide ATS optimization suggestions
• Add support for exporting tailored CVs to different formats

Built With

• claude
• claude-api-(anthropic)
• lovable
• react
• tailwind-css
• typescript