It started with an email.
A friend from a previous job asked if I could help her tailor her resume for a specific position. She had the experience but wasn't sure how to present it for this particular role.
I said yes.
And then my brain did what it always does.
The First Version Lived in My Inbox
I didn't start with a dashboard.
I started with email.
I set up a folder in Gmail just for her, and I told JJ (my AI assistant) to watch it. When she emailed her resume and a job description into that folder, JJ would tailor it and email the result back.
If she wanted changes, she'd reply. JJ would adjust and send the next version.
It worked. It was just kind of awkward.
She might be working on two different resumes at one time, emailing JJ with requests for edits, and the emails back and forth got confusing fast. It wasn't smooth.
That clunky first version served its purpose, though. It showed me what she actually needed.
It led me to figure out how to make the next version flow.
Then I Built the Dashboard
I started building a way to tailor her resume using AI.
Claude would take her background, the job description, and a set of rules about how to present experience honestly and effectively.
It would generate a custom resume and cover letter, highlight which keywords it matched, and flag any gaps.
But once I had the resume tailoring working, I kept seeing more that could help.
She needed a way to track the jobs she was applying to.
She needed an easy way to download the resume as a Word doc or print it as a PDF without any special software.
She needed a place to tell the AI about her career history so it could get smarter over time.
This was more than uploading a resume. She'd describe what she actually did at every job, in her own words.
That gave the AI real context to work with. Every tailored version after that came back sharper.
So I built all of that.
A five-tab dashboard: job leads, AI resume tailoring with a live progress checklist, a chat interface for career coaching, a place to share her background and preferences, and a help section.
Then I thought, why not train an AI agent to scrape the internet for jobs and populate her leads tab automatically?
So I built that too.
Not every scraped lead was a fit. So I built an ATS screener into Claude Code that scores the fit against each job from one to ten.
Anything below a four, she never sees.
Every design decision was filtered through one question: would she know what to do here without me explaining it?
She isn't a tech person. If something wasn't immediately obvious, I redesigned it until it was.
Then Two More Friends Needed the Same Thing
Two other women from the same workplace were also job searching.
Instead of building three separate tools, I set it up so one codebase could serve all of them.
Each person got their own site with their own data, their own job leads, their own resume tailoring, their own chat history with the AI.
Adding a new person takes about fifteen minutes.
The AI learns each person individually. It remembers what kind of roles they prefer, what experience to emphasize, what to downplay.
Every interaction makes it more helpful for that specific person.
Built to Run Without Me in the Room
The onboarding was critical. They each had to be able to use it without me sitting next to them.
That's always my goal: build something that runs without me in the room.
I built an 18-slide guided tour every user walks through before they see their dashboard.
It shows exactly what each section does, with real screenshots from the live site.
It references things they'd recognize from working together. Warm and personal, not a generic software walkthrough.
That tour does the same job a session with me would do. Meet them where they are, show them what matters, and make them feel like the whole thing was built for them. Because it was.
From One Email to a Platform
The whole thing took about 36 hours from that first email to three women having their own personalized, AI-powered job search dashboards.
Tailored resumes, career coaching, preference learning, job tracking, and an onboarding experience that makes it all make sense.
That's how I work: someone has a problem, I solve it.
And then I keep seeing ways to make the solution better until it becomes something none of us expected.
Want to see it for yourself? The demo is open and running at pam-dashboard-seven.vercel.app. Jane Doe's sandbox is loaded with sample leads so you can click around without signing up.