Graph IP Development
Co-inventor on U.S. Provisional Patent Application No. 63/911,215, System and Method for a Computational Graph, filed November 4, 2025.
Kaiser Permanente
Prior to founding NodeFox, I focused on operations and strategy at Kaiser Permanente, covering process analysis, operational analytics, workflow improvement, and vendor oversight. This experience defined my approach to modern enterprise technologies.
Project Recap
Key Engagement
NodeFox
I co-founded NodeFox and work across company direction, product, enterprise discovery, systems design, implementation requirements, go-to-market, fundraising, and commercialization. NodeFox is building an environment for designing and running increasingly complex AI systems with clearer structure, execution, and control.
Bridging intent and final delivery.
The hardest part of enterprise AI is rarely proving that a model can perform a task. It is deciding what should be built around it - and getting that system to work inside an organization.
Planning
Pinpointing high-impact AI potential that syncs with established business goals and complex operational realities.
Design
Specifying the particular interfaces, workflows, and feedback cycles needed to convert model logic into viable tools.
Backend
Architecting the technical structure-data flows, guardrails, and integrations-that support reliable model execution.
Delivery Phase
Building the final solution to ensure it satisfies technical specs while integrating smoothly into daily user operations.
I also lead efforts in enterprise discovery, stakeholder alignment, market entry, fundraising, and growth strategy.
Studies
James Myhre · Working Paper · September 2026 · Zenodo · DOI: 10.5281/zenodo.22836398
Tracking Errors to Real-World Outcomes
When a model yields an error, what determines if it results in a negative business impact?
This methodology paper introduces a controlled simulation framework to isolate how system design and organizational handoffs either catch or amplify model error. It moves beyond model evaluation to system evaluation.
James Myhre · Working Paper · September 2026
The Authority of Artificial Outputs
At what point does an AI-driven suggestion evolve into an automated and binding corporate action?
A conceptual framework exploring the "authority gap"-the transition point where downstream users stop verifying outputs due to system trust, pressure, or UI design, effectively turning software into an autonomous decision-maker.
About
I bridge the gap between theoretical AI capabilities and building dependable systems for large scale use.
Following UCLA, I managed strategic operations at Kaiser Permanente prior to co-founding NodeFox. My work focuses on constructing reliable AI architectures, product engineering, and research that ensures models translate into measurable business impact.
UCLA
→
Kaiser Permanente
→
NodeFox
→
Current Role