Product Career Ops
A free, open-source toolkit for product leaders. It turns your career history into a private evidence bank, then works with an AI assistant to score roles, draft grounded application materials, and coach your growth.
Write the truth about your career once. Every application after that starts from strength.
01 · Score first
Paste a job link, a description, or even a screenshot. The system grades the role against your real profile and tells you whether it is worth your energy.
One pasted link became a scored, active opportunity the same morning. Fifteen years of healthcare product work lined up with the mandate.
02 · Packets, not cover letters
For high-conviction roles, one command assembles everything at once. Nothing is invented. Every claim maps back to your evidence bank.
The packet pulled three themes out of the evidence bank. They became the application answer, grounded in things that actually happened rather than generated filler.
evidence.yml.
03 · One ledger
Every opportunity, score, decision, and outcome lives in a spreadsheet the system maintains for you. Dated, auditable, and yours.
The submitted application was logged with a follow-up date one week out, and interview prep was already built.
| Role | Score | Status | Next |
|---|---|---|---|
| SignalCare AI · Dir. Product | 88 | Applied | Follow up in 1 wk |
| Northstar Health · VP Product | 84 | Active | Build packet |
| Kinetic Robotics · Sr. PM | 71 | Watch | Recheck in 2 wks |
| Atlas Cloud · Group PM | 58 | Archived | Reason logged |
04 · Grow on purpose
The same evidence bank powers weekly reviews, quarterly planning, and skill-gap analysis. Market feedback becomes your next deliberate bet instead of a vague anxiety.
/pco develop weekly captures fresh wins while they are vivid/pco develop quarterly updates your thesis and target listThe workflow
Six steps from a job posting to a logged outcome. Each one is a short conversation with your AI assistant.
Hand the assistant a link, a pasted description, or a screenshot. /pco search
The role is graded against your profile and evidence. 80 or above goes Active.
Read the scorecard like a decision memo: pursue, warm intro, or pass.
Materials, prep, and outreach for roles that clear the bar. /pco search packet
The system never submits anything for you. You edit, choose, and send.
Outcomes go back into the ledger as market feedback for the next round.
Your data
Everything the system knows about you lives in plain files you can open, read, and edit. No database, no account, nothing uploaded to anyone's server by this project.
profile.yml
Target roles, constraints, differentiators, and your career thesis.
resume.md
Written once, broadly. Tailored versions are derived, never invented.
evidence.yml
Achievements, hard decisions, tradeoffs, and metrics.
sources.yml
Companies, job boards, newsletters, and communities to scan.
product-career-ops.xlsx
Every opportunity, scorecard, decision, and review, dated and auditable.
private/
One folder on your computer, backed by a private repository only you can see.
Privacy by architecture
Tools are public so anyone can use and improve them. Your career data lives somewhere only you can reach. The split makes mixing them structurally impossible, not a matter of being careful.
Getting started
Three things to have ready, five steps to run. Read the blue box before touching the Terminal.
You do not have to learn these commands. Open your AI assistant, point it at this project, and say: "Set up Product Career Ops for me, then help me replace the demo data with my real resume." The assistant runs every step below on your behalf and asks when it needs something from you.
This downloads the public toolkit to your computer.
git clone https://github.com/vmittal2690/product-career-ops.git cd product-career-ops
python3 -m venv .venv source .venv/bin/activate pip install -e ".[dev]"
Run the workflow on fictional data first, so you know what good output looks like.
pco demo-init pco opportunity add --input demo/opportunities/northstar-health.json pco opportunity list
Create a private GitHub repository for your data, connect it as the private/ folder, and replace the demo files with your real ones.
git clone https://github.com/<you>/product-career-ops-data.git private
pco reset --confirm RESET
pco doctor # checks everything is wired up
Pick one role that genuinely excites you. The goal is not a perfect application; it is an honest read on whether your evidence is strong enough.
/pco search <job link or pasted description>
The operating rhythm
The system compounds through a light weekly routine, not marathon sessions.
Refresh sources and review new roles that scored well. /pco search scan
Build packets for the top one or two roles only. Depth beats volume.
Send the applications and warm outreach you prepared, in your own voice.
Record outcomes and capture fresh evidence while it is vivid. /pco develop weekly
Once a quarter, /pco develop quarterly steps back further: which mandates keep fitting, which gaps keep repeating, and what the next deliberate bet should be. That is when the job search stops being a scramble and becomes a strategy.