tec.box builds tools for professional software engineers — consultants, solo devs, and small firms. Its first tool targets a problem every one of those people already has: tailoring a resume to a specific job description without rewriting it from scratch, and without an AI tool inventing experience that isn't there.
The core flow
A user keeps one or more "base" resumes as markdown — one per career track, for example. Given a job description, the app picks the best-matching base resume, explains why, and lets the user confirm or override that pick before generating a targeted version. Facts carry through unchanged: employers, titles, and dates never move. Only framing and emphasis do. The targeted resume, and any base resume, exports to PDF or DOCX in a template that matches the on-screen preview exactly.
Getting started doesn't require typing markdown from day one — an existing resume can be imported from PDF or DOCX and parsed into the same schema. Plain export is free, capped at three documents a month; AI-driven targeting sits behind a Stripe subscription.
What shipped
- Google sign-in and account management (Auth.js)
- Base and targeted resume storage on Postgres (Drizzle ORM, Neon)
- AI-driven job-description matching and resume targeting (Vercel AI SDK, Claude)
- PDF and DOCX export with a true WYSIWYG preview
- AI-assisted resume import from an existing PDF or DOCX
- Stripe subscription billing, gating AI features behind the free tier
- A production deploy on Vercel, live today
The timeline
The design spec was approved on September 12. By October 6, the app had shipped through 93 merged pull requests — this wasn't a single feature dropped and walked away from; it's a product that kept shipping fixes and refinements after its core flow went live. That's the shape of a Liftoff engagement: one real user flow, built on production-grade foundations (auth, database, billing), live in weeks rather than quarters.