Personal application copilot

Compare one profile against one job, then show exactly what to improve.

Start with evidence-based fit analysis, a clear radar view, and targeted CV adaptation. Keep manual review in the loop. Add automation later.

Open App Scaffold
  • Web-first, mobile-friendly for iPhone and Android browsers
  • Generic prompt foundation plus user-specific personalization later
  • Manual review remains mandatory before any application output

MVP

Deliberate, trustworthy, and easy to ship

1:1 Comparison

One uploaded CV or profile against one pasted job description or job link.

Evidence-Based Match

Must, should, and could mapping tied to explicit profile evidence.

Visual Gap Report

Radar diagram, strengths, gaps, and risk areas in one mobile-friendly view.

CV Adaptation

Concrete rewrite suggestions and missing-info prompts before export.

Workflow

What the first release should actually do

01

Ingest profile

Upload a CV or paste structured profile text.

02

Ingest job

Paste a job description or a job URL for analysis.

03

Score fit

Show requirement coverage, hard gaps, and explainable scoring.

04

Adapt CV

Generate targeted edits and route everything through manual review.

Primary Screen

The comparison report is the product center of gravity

What the report needs to do

One screen should explain the score, show the overlap visually, surface the real blockers, and convert analysis into concrete CV edits without forcing the user to read walls of AI text.

  • Top summary with fit score and recommendation
  • Radar comparison of profile versus target role
  • Must, should, and could requirement evidence table
  • Gap analysis tied to impact and suggested next actions
Comparison / Report

Evidence first, rewriting second

The report screen becomes the bridge between diagnosis and tailored CV adaptation.

Open the app workspace

Later Vision

Automation comes after trust and fit clarity

Profile Sync

Add LinkedIn authentication later so the user profile can be imported and refreshed automatically with clear consent controls.

Job Polling

Support saved searches, source connectors, and scheduled polling for LinkedIn, StepStone, and similar platforms.

Email Scanning

Optionally scan incoming emails for job links and score them against the stored profile.

Notifications

Alert users when a strong match appears or when recurring qualification gaps show a profile weakness worth fixing.

Prompt Strategy

Use the GitHub prompts as two separate layers

Generic System Prompt

The generic fit-check prompt becomes the reusable product default for scoring, mapping, and recommendation quality.

Personalized Profile Prompt

The personalized prompt becomes a user template for a saved profile, a preferred document style, and deeper application support.