Jonathan Barth
Technical Executive | AI Architect | Patent Holder
From military operations to fintech at scale. Principal Architect at PayPal. Technical Director at Schwab. 38 years of building, securing, and scaling enterprise systems — now building Orreth, a kernel for fleets of AI agents — identity, memory, rails, meter, gate, watches and levers held in one place under one law — and available for AI architecture leadership — fractional or full-time.
Years Experience
US Patents
Platform Uptime
Orreth Kernel Conformance Cases
The Flagship
Orreth — a kernel for fleets of agents
The systemic problem in agent fleets is entropy: every agent brings its own identity story, its own memory, its own way of calling a tool, its own idea of what it may do and what it costs. Orreth takes the rigid parts away from the agents and holds them in one small Rust kernel under one law — identity, memory, rails, meter, gate, watches and levers — while policy stays with people and an intention is the unit of control. Everything a body does lands on one signed record, and a person can always stop what the machine manages. Built solo, in the open, twice: a working first architecture was halted in September 2026 because the instruments said it was built for the wrong kind of growth, and this kernel is what replaced it.
AI Architecture & Strategy
The layer under the agents
Frameworks add loops; they do not remove the entropy. Put twenty agents on one problem and nobody can say who did what, under which law, for how much, or how to stop it. An operating system removed that entropy for programs by holding the rigid parts in one place under one law — Orreth does that for agents. Twenty-five years of identity, risk, and observability at PayPal and Schwab, applied to the kernel agentic AI is missing.
Identity & the Gate
A keypair is a self, and a self survives the process. An agent that dies comes back as the same self; a person sits with a seat the kernel minted after a proof. Joining is a governed request — a person's word admits a body, scopes only narrow, and refusal wears one face.
One Signed Record
Logs are not institutional memory. Every ask, reply, fact and tool call lands on one signed, append-only record — corrections are siblings, never overwrites; a purge is a governed act that leaves a hash. Every view is a projection over that one truth, and every thought lands in dollars beside the work.
Watches & Levers
The kernel watches its own doors, benches, leases and outbox. Every remedy is a lever a person can see; a routine one the kernel pulls itself and says so; a consequential one holds until a person says yes. The human can always stop what the machine manages.
Exploring senior agentic-infrastructure & AI architecture roles — the résumé and the proof-of-work.
Professional Services
Fortune 100 architecture discipline, available for your organization
Featured Projects
Showcasing key architectural achievements
Orreth
A kernel for fleets of AI agents
The flagship — one Rust kernel holds the rigid control every agent used to carry alone: identity as a keypair that survives the process, one signed append-only record, one envelope over the same rails, every thought metered in dollars, a five-status desk a person's word alone opens, and levers a human always holds. Policy stays with people. THE PANEL draws the whole world live at demo.orreth.ai.
0.1.0
Kernel
730
Cases
10
Crew
Commander.ai
Mission Control for your AI team
Multi-agent orchestration platform with 8 specialized agents, real-time prompt engineering, and performance analytics.
8
Agents
28
Endpoints
100%
Coverage
WorldMaker.ai
Enterprise Digital Lifecycle Intelligence
Digital lifecycle management platform with MTTD < 0, three-tier attribute system, and enterprise risk intelligence.
41
Endpoints
195
Tests
6
Languages
MCPFarm.ai
Kubernetes for AI Tools
Centralized MCP server orchestration with unified gateway, real-time dashboard, tool-level analytics, and Python SDK.
5
Servers
SSE
Gateway
Python
SDK
ProblemSolver.ai
Five-Agent Reasoning Pipeline
Structured reasoning pipeline with autonomous revision loops — Thinker, Planner, Critic, Solver, Judge — 15 specialized tools.
5
Agents
15
Tools
3
Models
ml-pipeline.ai
Autonomous ML — Raw Data to Trained Model
Self-improving ML pipeline with LLM-powered Critic loop, 6 specialist nodes, Optuna tuning, and Neural Observatory dashboard.
6
Phases
3+
Models
3
Loops