Good work.

    .    _    .  ,   .           .
    .  *  / \_ *  / \_      _  *        *   /\'__        *
    .    /    \  /    \,   ((        .    _/  /  \  *'.
   ..   /\/\  /\/ :' __ \_  `          _^/  ^/    `--.
    .  /    \/  \  _/  \-'\      *    /.' ^_   \_   .'\  *
   . /\  .-   `. \/     \ /==~=-=~=-=-;.  _/ \ -. `_/   \
  . /  `-.__ ^   / .-'.--\ =-=~_=-=~=^/  _ `--./ .-'  `-
  ./jgs     `.  / /       `.~-^=-=~=^=.-'      '-._ `._
  

Forward-deployed data scientist. 15+ years shipping ML inside messy, regulated environments - automotive, mineral rights, health care, maritime security, logistics. I go on-site (or virtually), dig up the data nobody mapped, figure out what the right question is, then build and deploy. Part Archeologist, part anthropologist, and part data scientist.

What I've done:

   - Deployed an AI validation pipeline for BIM/CAD against 400-page ESG standards, cutting review from days to minutes.
   - Built and integrated a part-failure prediction system into a German automaker's production workflow, avoiding lemon law recalls ($20M+ exposure)
   - Oil & gas well and lease production forecasting. 
   - Stood up real-time emissions detection on industrial smokestacks, preventing $75K-per-incident fines
   - Revenue optimization and persona identification for debt collections (15-20% net revenue lift)
   - Built an LLM-based legal document extraction system (land runsheets), collapsing 3-day DD cycles to 30 minutes.
   - Built vessel piracy risk engine that informed naval escort deployment, contributing to fewer hijackings.

   - Gambling.
   - Ads/Surveillance.
   - Payday loans/rent-to-own.

A warning: Don't burn Chrome.

I can finally have my own Dixie flatline. Cool.

It certainly increases shareholder value.

Email in profile.

The default regime is instability. Computational capacity is unstable through time, and the problem size itself changes (fractally) through time.

Most of what people call "data science" is now doable with an LLM and a weekend. I focus on the messy, ambiguous, regulation-heavy gnarly problems where nobody is sure what the right question is. Often it involves getting hands on to find the data and signal that's really needed.

What I've done:

   - AI verification of building and architectural BIM/CAD against 400 page ESG standards. (Days of work down to minutes)
   - Part failure prediciton, saving a German automaker from lemon law recalls. ($20M+ exposure avoided)
   - Oil & gas well and lease production forecasting. 
   - Real-time emissions detection for industrial smokestacks. ($75K fines per incident prevented)
   - Revenue optimization and persona identification for debt collections (15-20% net revenue lift)
   - LLM-based legal document extraction (land runsheet), reducing due diligence turnaround from 3 days to 30 minutes
   - Built vessel piracy risk engine that informed naval escort deployment, contributing to fewer hijackings

   - Gambling.
   - Ads/Surveillance.
   - Payday loans/rent-to-own.

One of those ships crashes and the boys from the Dwarf find the service mechanoid, which is how they get Kryten.

Naturally it got shut down and your media no longer plays.

Wolfram had come up with some normal techie names "computron", "math-o-matic" or whatever. SJ said No those suck, use something simple like Mathematica.

SJ recommended some of the UI bits of the notebook. Particularly the separators between cells.

The idea is decompile something like Wordperfect or Framemaker, then port the NeXTStep code to GNUStep and have WP on GNUStep/Linux.

We know Sick Boy (Zero Cool) would be an emacs user.