AI Ops Lead at Supermix
Day-to-day AI operations — managing LLM systems, shipping AI products, bridging the gap between research and production. Increasingly doing the building myself.
AI Ops Lead · Builder · MSEH
Harrison Dudley-Rode
AI-native builder. Published researcher. 52 apps shipped. Looking for the right team to build with.
Vancouver, BC•open to opportunities
Current work
What I spend time on
Building in public
52 production apps shipped in 4 months: AI workflow demos, voice AI assistant, multi-agent systems, context engineering studio, AI interview simulator, salary negotiation simulator, behavioral interview coach, system design interviewer, legal AI pipelines, autonomous research agents, code migration tool, prompt evaluation tools, workout tracker, finance dashboard, MCP server demos. All live, all in the portfolio. Learning by doing — fast.
Research Associate at SPRINZ
Published researcher on exercise physiology and endurance performance (AUT, New Zealand). The science informs the apps — Durability Analyzer was built directly from my own peer-reviewed methodology.
Selected work
A smaller public set
These are the projects that best represent the mix of AI systems, product judgment, and sports science work on the site.
beef
Built in 4 weeks
A workout tracker that actually understands strength training
Next.jsSupabaseTypeScriptSports Science
TrainingLoad
Built in 1 week
Evidence-based training load monitoring for endurance athletes
Next.jsTypeScriptSports ScienceRecharts
Pipeline Demo
Built in One overnight session
A 4-stage LLM pipeline where you can see every step
Next.jsAnthropic ClaudeLLM OrchestrationTypeScript
AI Research Agent
Built in One overnight session (March 8, 2026)
AI Research Agent — Built with AI assistance in one session
Next.jsAnthropic ClaudeSSE StreamingAI Agents
Live dashboard
Agent activity
A live snapshot of the repos, pull requests, and shipping cadence behind the systems I build with AI agents.
Further work
Projects and research
Projects
A smaller set of product, data, and sports science projects that best represents how I work.
View projects →Contact
Interested in working together?
Email is best. GitHub and LinkedIn are there if you want to see more of the work behind the page.