PM Tooling · Built with Lovable

ApplyFlow — AI Job-Search Command Center

An AI-powered job-search command center built specifically for Product Management roles. ApplyFlow tracks every application, discovers relevant openings, scores each job against my profile, and manages follow-ups — replacing a scattered spreadsheet with one dashboard, for my own job search.

Role
Solo builder — problem to working prototype
Method
Workflow design, fit-scoring logic, prompt engineering
Scope
Personal PM job search, used daily
The problem

A job search run from five different tabs is a job search you lose track of

Spreadsheets, job boards, email threads and calendar reminders don't talk to each other — so applications get missed, follow-ups slip, and there's no single view of where things actually stand. I built ApplyFlow to run my own PM job search from one place instead.

Solution

One dashboard: track, discover, score, follow up

TRACKING

Every application, one pipeline

Applications move through a clear pipeline instead of living across scattered spreadsheet rows and inbox threads — status, next step and last touch always visible.

DISCOVERY

Relevant openings surfaced, not searched for

Instead of re-running the same searches across job boards, ApplyFlow surfaces openings that match the roles I'm actually targeting.

SCORING

Fit-scored against my real profile

Each opening is scored against my own experience and target criteria, so time goes to the roles worth a tailored application, not every listing that matches a keyword.

Prototype

A working command center, not a spreadsheet template

ApplyFlow is built and live — application tracking, job discovery, fit scoring and follow-up management all work end to end in the prototype below.

Live prototype — ApplyFlow Open full screen ↗
Open interactive prototype ↗
Process

Built to solve my own problem first

Discovery
Problem discovery — started from my own PM job search falling out of sync across five different tools.
Flow
Workflow design — mapped a single pipeline for tracking, discovery, scoring and follow-ups.
Scoring
Fit-scoring logic — defined the criteria and prompts that score an opening against my actual profile.
Build
Prototype — shipped a working, clickable command center on Lovable and used it for my own search.