87n1
87 in 1. That's the ratio, of orchestrated AI workflows to human input. The future of building is a team, and the team includes agents.
Jeff Michael Johnson
Full-Stack Developer & AI Workflow Architect
Design-Driven Full-Stack Engineer
I'm aFull-stack engineer,
shipping iterative agent systems.
I build multi-agent systems for my own work, then extract the reusable patterns into open-source tooling and shipped products. Currently operating across mobile, web, and native platforms with a focus on orchestration infrastructure that holds up under real constraints.
Command Deck · Actual Screen Recordings
My workstation ships front end + back end, iterates, improves, and keeps shipping. Human agency + AI for the win. I put in serious work to make sure my training on new AI tools is ahead of the pack.
I'm a design-driven full-stack engineer and AI workflow architect building production systems across mobile, web, and native Apple & Android platforms. My work focuses on two outcomes: shipping polished user experiences for real people, and removing the coordination overhead that slows delivery, by designing orchestration, hooks, memory, and automation that make development compound over time.
The design comes first. I work in Figma, define a visual language and motion system, then translate that into production code. PrayerMap was born this way: the "Ethereal Glass" design system shipped as a formal handoff package in November 2025, translated into a React web app, then migrated to React Native with a 470-line animation preservation spec. See the design case study.
Recent work includes PrayerMap: a real-time, map-based prayer platform built with React Native and Expo, backed by a Supabase/Postgres stack, and supported by a six-dashboard operations command center for moderation, outreach, and behavior monitoring. In parallel, I’ve open-sourced infrastructure like Neurotoken, a test-backed prompt scoring engine that runs in Claude Code hooks, along with tooling patterns that keep multi-agent work reliable rather than chaotic.
Before software, I worked in institutional finance and later operated businesses. That experience shaped how I build: clarify risk, instrument reality, respect users, and ship with discipline. Today I'm based in the Phoenix area, and open to full-time roles, contract work, and high-signal collaborations.
What I'm building right now
Three projects in flight. Not shipped yet, and labeled that way on purpose. Here is where each one stands and where it is headed.
Portfolio
Major Projects
18 projects across mobile, web, AI, and automation
PrayerMap
Making the Invisible, Visible.
A full-stack mobile app connecting communities through interactive, real-time prayer mapping. Features a 6-dashboard command center with AI behavior monitoring, CRM pipeline, and RAG training controls.
Puber
Built for the long miles.
An ultramarathon support app connecting 100-mile trail runners with pacers and crew, plus a per-race coordination board for rides, lodging, drop bags, and last-minute replacements. Expo + Supabase + a Next.js ops console: 25 migrations with RLS on every table, 14 edge functions, a Postgres match state machine, an offline Race Kit, and safety designed in as architecture. Built end to end and running locally; not yet deployed. Target: iOS + Android for the 2026 100-mile season.
GBuild
Build anything. Smarter.
A macOS desktop app that orchestrates 12+ AI coding agents (Claude Code, OpenClaw, Hermes) from one calm, voice-first interface. Electron + a Rust sidecar over JSON-RPC, pluggable agent drivers, a whisper/Piper voice pipeline, a RAG layer, bidirectional Obsidian sync, an MCP gateway, and the "G" mascot. Marketing site live at gbuild.app; the desktop app is in private development.
Johnson Power Rankings
A world ranking for the 200+ mile ultra.
A data-driven, weekly world ranking system for ultramarathon runners, launching with the 200+ mile tier. A pure-TypeScript engine (course-difficulty model, iterative field strength, time-decayed weighted-mean ratings, backyard scoring) over a Supabase backend with 11 migrations, RLS on 20+ tables, 6 edge functions, and a Monday recompute cron. A Next.js web app, a separate admin panel, and an Expo mobile app. Built with my dad as the domain expert. Live in preview; public launch pending.
Claude Code Toolkit
Production infrastructure for AI agents.
An open-source infrastructure layer for Claude Code with 13 interactive skills, 7 specialized agents, a Smart MCP Loader that auto-enables servers via NLP scoring, and 48 scaffolding templates. Built entirely with Claude Code.
Neurotoken
Adaptive thinking allocation.
A zero-dependency prompt scoring engine that classifies every Claude Code prompt on a complexity × stakes matrix, recommending one of 11 model/effort tiers from haiku/low to opus/max. Adversarial-tested with 183 passing tests.
Directors Chair
Mission control for an autonomous agent fleet.
A real-time mission control dashboard monitoring an autonomous agent fleet across job search, project maintenance, and infrastructure.
Mother Hen
Mission control framework.
The underlying framework powering Directors Chair. Provides agent lifecycle management, pipeline orchestration, persistent state, and multi-operator support. Built with Next.js, Supabase, Qdrant, and Discord. Private beta, with open-source release planned.
LivCut
Privacy-first video editing.
A professional video editor built with Swift, Metal, and CoreML: 66 Metal compute kernels and 11 on-device AI engines. All AI processing runs locally on the Apple Neural Engine.
Challenge Cup
Ryder Cup for your crew.
A multi-platform golf tournament app with AI-powered caddie, ESPN-style broadcast leaderboards, and format-per-segment flexibility. Built for Apple Watch, Garmin, and mobile.
Living Earth
The planet, alive.
A WebGPU-powered globe visualization rendering live aircraft, satellites, ships, weather, and prayer energy at 60fps. GPU-computed boid flocks and particle simulations on mobile via react-native-wgpu.
MCP-WP
AI Meets WordPress.
A LangGraph-based architectural prototype for multi-agent WordPress analysis. Orchestration harness and realtime Cursor-to-Pinecone sync are live; 27 agent personas defined in Postgres.
Ora
Multi-agent orchestration, extracted from production.
An AI orchestration framework extracted from the production MCP-WP system. Coordinates Claude Code, Cursor, GitHub, Pinecone, LangSmith, and Slack into a unified development pipeline with file-based agent coordination.
InkProof
Print preflight for Canva.
A Canva app that catches print production issues before they cost money. Checks safe zones, image resolution, CMYK safety, and dimensions. Built and submitted to the Canva Marketplace; currently under review.
Rapid Studio
AI-Powered Creation.
A GPU fleet orchestrator for rapid AI content generation. Microservices architecture with FastAPI orchestrator, Redis queue, and RunPod GPU workers generating images in under 1 second.
n8n Media Workflow
Autonomous content engine.
An autonomous Instagram content system that analyzes accounts, extracts style patterns, generates brand-aware content at 5 posts/hour, and learns from human feedback via Slack.
MeetingCost
See what meetings really cost.
A macOS menu bar app that tracks the live dollar cost of any meeting. Swift 6 + SwiftUI, StoreKit 2 IAP, App Sandbox, zero network. Built end-to-end by a Claude Code app-builder agent from a SPEC.md file — 11 days from scaffold to the Mac App Store.
PasteDrop
Smart clipboard transforms.
A macOS menu bar app that detects what you copy — JSON, URLs, hex colors, emails, code, UUIDs — and offers one-click transformations tailored to that content type. Native, fully offline, Keychain-sandboxed.
Workflow
The Orchestration Layer
How I build with AI-accelerated workflows
User Intent
Natural-language requirement or idea
AI Director
Ora orchestrates agents & context
Claude Code + Cursor
Parallel implementation agents
GitHub
Automated commits & CI pipeline
Deployed
Live in production, zero-touch
11
Shipped Projects
Web, mobile, native iOS, and AI systems
6
AWS Courses
3 tracks · 18 days · Oct–Nov 2025
29
n8n Workflows
14 active · 261 nodes · fully automated
Live Production
n8n Automation Workflows
Real production workflows running on n8n Cloud. Each graph below is built from live workflow data: click to expand and explore the node topology.
29
Workflows
14
Active
261
Total Nodes
6
Integrations
The Orchestration Layer
Designing complex multi-agent workflows where AI agents coordinate, reason, and execute in parallel — with tools, memory, and feedback loops.
AI Constellation
The assembly line changed manufacturing. AI agent pipelines are the next one — 16 technologies in orbit, everything connected.
Powered by AI · Hover a node to explore
Some of this is family
Two of the things I'm building right now came out of the same world: the ultramarathon one. My dad, Paul, is an ultrarunner, a pacer, and a crew member. He knows the runners by name, the courses by feel, and the culture from the inside. Five consecutive Cocodona 250 finishes, a 1,000-mile buckle, and the Arizona Monster 300.
Johnson Power Rankings started at the kitchen table, on a recording: Dad as the domain expert, me on the stack. There's no credible data-driven ranking for the 200-plus-mile races, so we built one. The name is the family name, on purpose.
Puber is the other side of it: an app that connects 100-mile runners with the pacers and crew who get them through the night, plus a per-race board for everything that falls apart at the last minute. It is built end to end. It is not shipped yet, and I'm saying so.
He doesn't run alone. My mom crews every race. My nieces and nephew train at IMG Academy. The whole house leans into hard things.
