How I built a privacy-first resume builder with $0/month server costs (Now launching!)
Hey everyone,
Like many of you, I got tired of seeing everyday utility tools hidden behind predatory paywalls and data-harvesting sign-up walls. Applying for a job shouldn't require giving away your personal data to another database.
That's why I built ForgCV — a 100% free, no-signup resume builder.
To keep long-term maintenance costs at exactly $0, I designed it with a pure client-side architecture:
• Tech Stack: Next.js + Tailwind CSS.
• Hosting: ForgCV is currently hosted on Netlify
• Data Storage: LocalSpac/LocalStorage (Data never leaves the user's device).
• PDF Rendering: Done entirely client-side, offloading the heavy lifting to the browser.
The result is a fast, vector-PDF output with clean HTML/CSS semantics that parse perfectly on ATS systems like Workday and Greenhouse. It also supports 8 languages with full RTL support.
We are officially live on Product Hunt today! I'd love to get the community's feedback on the client-side approach and the overall UX.
Check us out here and let me know what features I should build next: [https://www.producthunt.com/posts/free-cv-builder-in-2-minutes]
nice
thank yousi much
Love this. From the hiring side, I can tell you ATS parsing trips people up way more than they'd expect, and it's almost never the content, it's the layout. Multi-column resumes and skills tables especially tend to get scrambled by Workday/Greenhouse even when they look clean. Curious if you've stress-tested ForgCV against those? If it holds up, that's honestly a bigger selling point than the $0 hosting.
Coming from the hiring side, your confirmation means the world. You are 100% right—I see brilliant candidates get filtered out constantly not because of their talent, but because a multi-column template or a fancy skills table turned their experience into unreadable garbage inside Workday.
To answer your question: yes, stress-testing against this exact "scrambling" effect has been my main focus.
Since I engineered ForgCV with a strict single-column, linear reading order and clean semantic HTML headings, the text flow stays completely intact. I’ve been running the generated PDFs through actual raw text extractors and parser simulators to ensure that what Greenhouse "sees" is a perfectly ordered, clean text stream—no overlapping text, no broken tables.
You made a fantastic point. For a job seeker, saving $20 is nice, but knowing for a fact that their resume won't get shredded by an automated portal is the real game-changer.
I’m currently working on documenting these stress tests to show the hard proof right on the landing page, so applicants can apply with total confidence.
Thanks for bringing the hiring manager's perspective into this thread!