GTM Engineer|Content & Brand Strategist|Automation & Software Engineer
LOGANMOSS
Marketing & machine logic, powered by caffeine, automation, APIs, and a whole lot of Airtable tabs too.Side quest enthusiast.
- $100K+
- revenue generatedLoganRDJ LLC — brand strategy consultancy
- 10,000+
- concurrent viewersLivestreamed content, peak concurrent audience
- 30%
- manual process cutAutodesk — webhook-driven lead pipelines
- 100+
- engineers mentoredTrilogy Education — full-stack bootcamp
Selected work
I make the thing.
Then I make it run itself.
Software & Automation
Systems that keep running when nobody is watching them.

Spacious
macOS menu bar app for multi-monitor setups. Split each display into a grid of up to 48×48, drag out zones, and assign apps to them — left half of the widescreen to Chrome, the vertical monitor to Slack over Terminal — then apply the layout in one click. Saved layouts switch by hotkey, any window snaps with ⌃⌥Space, and grids are remembered for monitors that get unplugged. Swift and SwiftUI with the macOS Accessibility API for window control, grid maths isolated in a tested SwiftPM package, projects generated by XcodeGen. Runs entirely on-device: no account, no backend, no network.

Homebound
AI-powered lead prioritization and outreach assistant for real estate agents — built to surface who to call next instead of burying it in a spreadsheet. Syncs leads from Follow Up Boss and ranks them with a hybrid rules + LLM scoring engine that explains every verdict in plain language, then matches an agent's own listings against each lead's stated criteria. FastAPI, PostgreSQL, SQLAlchemy and the OpenAI API on the backend; Next.js, TypeScript and Tailwind on the front; plus an offline-capable PWA for taking notes in the field.

FridgeBook
Kitchen inventory app that tracks what's in your fridge and pantry, flags what's about to expire, and keeps a shopping list that puts groceries straight into your kitchen once they're bought. A ground-up rebuild of my original bootcamp project. React, React Router and FullCalendar on the front end behind an animated open-fridge landing page; an Express and Sequelize API over PostgreSQL with Umzug-versioned migrations; Google sign-in through Passport OAuth 2.0 with per-user data and Postgres-backed sessions; 100+ automated tests running in GitHub Actions on every push. API on Render, front end on Netlify.
Creative Direction
Campaigns and shoots for brands people already know.

100Thieves x Crazy Raccoon x LoganRDJ
Merchandise shoot for 100Thieves with Tokyo-based Crazy Raccoon clothing brand.

Slander - Thrive Collection
EDM DJ and Producer Slander, Thrive Merch line in collaboration with @supcolleen

Slander - Eye Swimsuit Collection
EDM DJ and Producer Slander, Eye Swimsuit Merch line in collaboration with @karinerino
Stack
What I build with.
- Languages
- PythonSQLJavaScriptTypeScriptJavaSwift
- Frontend
- ReactNext.jsReact NativeSwiftUIGatsbyJSHTML5CSS3Tailwind CSSMaterial UIBootstrapMapbox GLPWA
- Backend
- Node.jsExpress.jsFastAPIFlaskGraphQLREST APIsPostgreSQLMongoDBMySQLSQLAlchemyAlembicPydanticOAuth 2.0JWT Auth
- Developer Tools
- GitGitHubDockerVitepytestJupyter NotebookPostmanNetlifyRenderClaudeGPTGitHub Copilot
- Platforms & Automation
- OpenAI APIGmail APISalesforceMarketoHubSpotZapierAirtableGoogle AnalyticsTableau
Proof
Every number has a
system behind it.
LoganRDJ LLC — brand strategy consultancy
- Input
Brands with an audience they couldn't convert, and creators with reach but no commercial structure behind it.
- Built
A consultancy practice pairing campaign strategy with the tracking and automation to prove it worked — sponsorship pipelines, partnership terms, attribution.
- Output
Over $100K in revenue across partners including Taco Bell and AT&T, with campaigns measurable enough to renew on.
Build notes
Working in the open.
Why I left Hermes: multi-agent is a protocol problem, not a model problem
Hermes gave every agent a great brain and no reliable way to talk to the others. Here's what broke, what I evaluated instead, and why I still haven't fully committed.
Building Homebound: scoring real estate leads with rules and an LLM
Why a hybrid rules + LLM engine beat either approach alone, and how every score ended up explaining itself in plain language.