System 01
Positioning Engine
Turns senior experience into a crisp market thesis: target roles, salary logic, proof points, constraints, and non-negotiables.
romanmartins.com
Personal lab · build notes · documented in the open
AI Job SearchOS
The workspace I built to run my own senior job search: positioning, role scoring, tailored applications, proof-of-work artifacts, outreach, and interview prep – one command center, run with Claude Code.
- Shape
- Six systems
- Format
- Personal lab
- Status
- Running daily
- Next
- Build notes
Claude CodeCodexMarkdownGitRole scoringCV tailoringOutreach CRMInterview briefs
The problem
Senior job searches are execution problems now.
Better roles rarely come from more generic applications. They come from sharper positioning, better target selection, proof that travels, and a rhythm you can run every week. This page documents the operating system I built for that work – and what it produces in a live search.
job-search-os/current-data.md
/screen-jobs score remote AI PM roles
/tailor-cv map evidence to one JD
/proof-artifact draft a 30/60/90 AI roadmap
/outreach build warm contact paths
/interview-prep turn notes into answer briefs
The system
Six systems, one workspace
Each system produces working workspace assets – trackers, rubrics, drafts, artifacts – not documents about work.
System 02
Role Radar
A role-screening pipeline that finds, filters, and scores opportunities against fit, leverage, compensation, and remote reality – then verifies they are actually open.
System 03
Application Factory
Repeatable Claude Code workflows for tailored CVs, cover notes, recruiter messages, and application decisions – selecting and rewording true facts, never inventing.
System 04
Proof-of-Work Lab
The artifact that makes a candidate easier to hire: AI roadmaps, teardown memos, product audits, and live tools built for one specific company.
System 05
Outreach System
A contact map, artifact-led messages, follow-up cadence, and weekly execution loop that does not depend on willpower.
System 06
Interview Room
Converts the workspace into prep packets: company research, story bank, decision criteria, compensation notes, and closing questions.
What the workspace holds
- A clear target-role thesis and positioning stack
- A working job-search workspace with commands, trackers, and scoring rubrics
- A reusable CV and application tailoring flow
- Proof-of-work artifacts built per target company and shipped live
- An outreach rhythm with contact research, scripts, and follow-up loops
- Interview prep briefs that turn messy notes into sharp stories
How it runs week to week
A morning screening pass that finds, verifies, and scores new roles against a locked rubric.
A tailoring flow that maps real evidence to one JD – selecting and rewording true facts, never inventing.
Proof-of-work artifacts built for specific target companies and shipped as live pages and tools.
An outreach rhythm with named contacts, artifact-led drafts, and follow-up loops – logged, not remembered.
Who tends to find this useful
- Senior PMs, product leads, and operators moving into AI-heavy roles
- Engineers and technical consultants who want a sharper commercial narrative
- Experienced professionals targeting remote, high-salary, or international roles
- People who want to build a personal system instead of collecting templates
Not for
- You want a one-click resume generator
- You are looking for a course to buy today
- You prefer generic AI resume tools over building a personal system
FAQ
Useful details
Is this a course?+
No. It is my working system, documented. I occasionally compare notes with people running similar searches, and the note form below is the quiet list that would hear first if a version of this ever opens up – but nothing is packaged or sold here.
Do I need to be technical to run something like this?+
You do not need to be an engineer, but the system assumes you are comfortable working in structured files and using AI tools. It is built for leverage, not as a passive template pack.
Does a system like this guarantee a job?+
No. The point is not a guarantee. The concrete value is a sharper job-search operating system: better positioning, better role selection, stronger artifacts, and a clearer weekly rhythm.
What tools does it use?+
Claude Code, Codex, Markdown, Git, and lightweight trackers. The system is built around repeatable AI workflows rather than one-off chat prompts.
Personal lab
A system I am running, not selling
I am not packaging this as a course. AI Job Search OS is the workspace behind my own search – documented here so the method is inspectable. If a future version ever opens to others, the note form below is the quiet list that hears first.
Interest
Share a quiet note
If this resonates, send a short note. Nothing is stored on the site. Your email client opens with a structured draft, and there is a copy fallback if that is easier.
Good people to compare notes with
Senior professionals navigating a real transition, moving toward AI-heavy roles, or trying to build a more systematic search without outsourcing their judgment.