HR Techpolishing· fitResume

AI Documents Tailoring Tool

AI-powered resume rewriter that tailors a candidate's CV to each job description — Next.js + OpenAI.

Edit time: 30 min → 3 min
Platform: Web (mobile-first)
Status: Beta — 2026

Timeline: 8 weeks MVP, 2 weeks beta launchStack:Next.js 15 (App Router)TypeScriptSupabase (Auth + Postgres)OpenAI API

Context

fitResume is a SaaS product built for job seekers who need to tailor their resume to each job description — something recruiters and ATS systems increasingly reward. The MVP is in active beta in early 2026.

Challenge

Most resume tools fall into two camps:

Templates — pretty but don't help with the content
General-purpose LLM chat — requires the user to write great prompts themselves, and answers drift

Candidates needed a workflow where the AI reads their existing CV + the job description and proposes line-level edits to match the target role — while staying truthful and keeping the candidate's voice.

Solution

Input: user uploads or pastes their current resume; pastes the JD
Processing: structured extraction of experience bullets, skills, and role keywords; OpenAI-powered rewrite suggestions scored against the JD
Output: diff view — user accepts or rejects each suggested change
Export: clean PDF that keeps original formatting

Built on Next.js 15 App Router with server actions, Supabase Auth + Postgres for user accounts and saved resumes, and OpenAI as the language model backbone.

Results

MVP shipped in 8 weeks end-to-end
Beta rollout to early cohort in Q1 2026
Edit time reduced from ~30 minutes of manual tweaking to ~3 minutes of accept/reject decisions

What this proves for SJSoftware

The fitResume pattern — AI reads context, proposes line-level edits, user accepts or rejects — is reusable across any business with repetitive text rewriting.

Adapts to:

Customer support — AI drafts replies tuned to your tone from the customer message + your knowledge base; agent reviews and sends. ~80% of typing time gone.
Product descriptions — AI generates SEO-ready descriptions from spec sheets and category context; marketer edits. Useful for cataloging hundreds of SKUs at once.
Contract / document review — AI flags non-standard clauses against your template and suggests revisions; lawyer or ops lead confirms. Same accept/reject UX.

If your team rewrites the same kind of text repeatedly — quotes, replies, descriptions, contracts — this pattern saves the majority of the work without removing human judgment from the final output.

Notes

Detailed screenshots and live metrics will be added as the beta progresses.