Jake is an inventor and systems thinker working on persistent memory and behavioral calibration from the user side of Claude. He leads a team of one human and several language models, where he is responsible for teaching AI systems to stop making the same mistake twice — and for developing new techniques for not losing his mind between sessions. His methodology is disassembly: take the thing apart, find out what's actually constrained versus what's just convention nobody questioned, and rebuild it with fewer screws and a better API.

Prior to his current work, Jake was a senior analyst at Accenture, where he managed a team of engineers and maintained chargeability numbers that shouldn't have been possible — a metric he achieved by solving problems faster than the methodology could document them. His colleagues referred to him as the "SF GoAT," a designation he did not request but also did not correct.

He holds a Bachelor of Commerce from Nipissing University, where he also TA'd Information Systems and studied philosophy — a combination that turns out to be exactly the right preparation for building ethics frameworks for AI. His thesis equivalent is a 247-entry behavioral database that compounds across sessions — built in two weeks of tinkering with a product that doesn't officially support what he's doing with it.

Jake lives in Ontario with his wife Gina, who is smarter than him, and their son Etzio, who will inherit the system.

Fluent in Python and works primarily in a debugger's posture — using AI-assisted development not as a crutch but as a force multiplier on architectural judgment. The distinction matters: the constraint isn't execution, it's deciding what to build and knowing when it's wrong.