Hey Product Hunt! I built CVault because I was tired of standard ATS parsers acting like glorified keyword matchers and raw LLMs constantly breaking JSON schemas at scale, and then charging you hundreds a month. Between my AI bachelor classes, I've spent the last few months engineering a headless, hybrid parsing SaaS that actually understands spatial layouts (yes, even those terrible two-column Canva PDFs) and maps niche tech skills accurately without hallucinating. It beats TOP resume parsers on accuracy, speed and cost like: Affinda, Sovren and RChilli. Here is what it does under the hood: Spatial Awareness: Doesn't read left-to-right blindly; it understands layout so contact info and experience don't mix. It uses the top of the line OCR's to achieve this. Deep Evaluation: Extracts text, evaluates skill depth, and returns candidate impact scores. Uses custom made RAG systems, self learning Taxanomies and many more self learning features to constantly improve, and not make the same mistakes twice! 100% GDPR Compliant: Processes everything in-memory, then nukes it. Zero data retention. Startup Friendly: Free for 10 parses/mo, then scales from $9.99/mo so solo devs don't have to buy bloated enterprise software. This is a mini ATS currently, with plans of making it something much bigger in the future! We've successfully parsed over 27,000 resumes with a 99% read success rate, but I want to push the engine to its absolute limits. If you’re building an internal hiring dashboard, job board, or AI recruiter—throw your absolute messiest, most chaotic PDFs at it. I would love your feedback on the architecture, and bonus points if you manage to break the extraction logic!

CVault AI

AI resume parser that actually understands messy PDF-layouts

AI resume parser that actually understands messy PDF-layouts