// Own project
AI Job Application Tool
Tauri desktop app, five Flask microservices, three-tier LLM fallback
/portfolio · 2025 / 2026
Hi, I am Joshua Mosley
AISolutionsEngineer
Applied AI·Implementation·AI Orchestrator
Sport scientist who builds production AI systems. Currently shipping for my own supplement brand suppstantial.com and for clients. Köln, Deutschland.
📍 Köln · DeutschlandOpen to roles
// 60 to 90s introduction · who I am, what I built, what I'm looking for
/projects · interactive showcase
Selected work.
Drag, swipe, or use the side arrows. Click the focused card for the full case study.
// Own project
Suppstantial Storefront
Custom Next.js commerce, Claude-powered assistant, creator dashboard
// Automation
AI Content Pipeline
Scrape, LLM analysis, multimodal generation, video render
// Own project
MojoMakes Portfolio + Client Portal
Custom workspace for client redesign projects
// Client project
Christian Hahn PT Website
Personal trainer redesign with AI chat
// Own project
Progressy
Native iOS + web goal tracker
/about
Background.
B.Sc. in Sportwissenschaften (Deutsche Sporthochschule Köln, 2019) and M.A. in Sportbusiness Management (IST-Hochschule, 2024). I co-founded Suppstantial in 2019, a supplement brand based in Köln, which is where I started writing software seriously.
Around 2024 I went deep into AI implementation. Today most of my work is full-stack with Claude, OpenAI, Cloudflare Workers, and Next.js, both for Suppstantial and for client projects.
17 years in sport and fitness gives me real domain context for health, behavior, and adherence problems. That tends to change what I build, not just how I build it.
I'm looking for AI implementation consulting, AI solutions engineering, or applied AI roles. Open to remote, Köln, or hybrid.
/approach
How I work.
01
AI as a starting point
I use LLMs, image models, and tools as part of the build, not as a final-step bolt-on. The point is to get to a working system faster, with the same level of judgment about correctness.
02
End-to-end
I work across the stack: schema, services, UI, deploy. Multi-LLM fallbacks, microservices, and webhooks come up in most of my projects. I prefer owning the full path over handing off in the middle.
03
Domain bridge
17 years in sport and fitness, plus running a DTC supplement brand. That tends to change what I build for health and behavior products, not just how I build it.
/stack
Stack.
// AI / LLMs
Claude APIOpenAI APIGeminiKimi K2WhisperNano Banana ProLangChainRAGPrompt Engineering
// Frontend
Next.js 16React 19TypeScriptTailwindFramer Motion
// Backend & Data
PythonFlaskFastAPINode.jsDrizzle ORMBetter-AuthSupabasePostgreSQLD1
// Infra & Runtime
Cloudflare WorkersOpenNextWranglerR2VercelTauri 2.0Playwright
// Workflow
Claude CodeMCPGitRemotion