Hi HN, I’m a frontend engineer and built Jobbi.app, a small tool that uses AI to tailor a resume to each job description. It’s currently free, unlimited, with PDF export. Problem During my own job searches I kept rewriting the same resume for dozens of roles. Most “AI resume builders” I tried either: – focused on generating a resume from scratch instead of working with my existing one – had strict free‑tier limits or aggressive upsells – produced very generic, obviously‑AI text. What Jobbi does – you upload a “master” resume (PDF or text) – you paste a job description – it extracts relevant parts, rewrites bullets to match the JD, and outputs a tailored resume you can edit in the browser and export as PDF. Under the hood it: – parses the resume into sections (experience, skills, etc.) – scores which bullets/skills are relevant to the JD keywords – rewrites only those parts, keeping the original structure and tone as much as possible – uses Gemini API with some prompt engineering around “don’t invent experience / don’t change dates / keep numbers”. What’s different – it assumes you already have a resume and just want fast tailoring, not a full CV builder – no limits / logins / credit cards right now (I care more about usage + feedback than monetization at this stage) – I’m optimizing specifically for software/tech roles, so I’d love feedback from this crowd. Questions for you – For those who’ve used similar tools, what did you hate the most? – Is there something obviously broken / insecure / unethical in my approach? – What features would make this actually useful for HN readers (e.g. integration with “Who’s Hiring”, diff view of changes, local‑only mode, OSS version, etc.)? Link: https://jobbi.app I’ll be in the thread to answer questions and hear any feedback (incl. “don’t do this because X”).