Working systems for job search, document filing, and personal finance.
This year I let hundreds of people go cold because I could not follow up with all of them. When I went back to the job search, I could not easily go back to my warm contacts, if at all. That is where this started.
I refused to apply at the spreadsheet like I did before. Hiring managers and applicants are both racing with AI now to get in as fast as possible, and I was not going to run that race by hand.
Hiring positions are a lot harder to find than the actual people, and my contacts did not have ins at the companies that were hiring. So I had to source and message people at those companies as fast as I could. How do you find them, track them, draft every message quickly, and still tailor each one? The Job Outreach Engine is my answer, and it took more than 13,000 lines of code.
Every system here started the same way, as something I was living through, built until it stopped being a problem. All of it is public code, linked from the Code page.
I use AI all the time to code. I come from operations, and I build working systems for problems I actually have. Each one here runs daily and the source is public, so you can check the work instead of taking my word for it.
Live from the Job Engine
These update themselves: the engine recomputes them every 15 minutes, and this page re-reads them at most every 5 minutes. A figure can be up to 20 minutes old.
AI-screened job-search pipeline with a live CRM, résumé compiler, and outcome tracking.
Automated document intake and filing for small law firms and RIAs.
Personal income and spending tracker, with monthly rollups and a category breakdown.
The Flask app serving this page, with a private read-only Console rolling up live metrics from each project’s own database.