From AI Curiosity to Everyday Craft: Closing the Adoption Gap (and Getting to Vibe Coding)
A draft proposal to close the AI adoption gap by standardizing agent workflows, improving observability, and enabling responsible "vibe coding".
Melbourne-based, building for global teams
I turn complex ideas intohigh-impact software
I am Paul Gradie, a software engineer, founder, and inventor with a PhD background. I help teams move from research to production by building foundational AI models, applied AI systems, and resilient software that keeps delivering under real-world pressure.
Latest Writing
View allGuild Driven Development: The Review Guild Model for AI enabled development
When code output is cheap, trust becomes the constraint. Guild Driven Development makes review the control surface that keeps systems coherent.
Seamless AI Agent Thread Transitions: A Debugging Success Story
How I successfully transitioned context between AI agents to solve a complex Sailfish test discovery issue, turning a long debugging session into a collaborative success.
Building Work Relationships
Building positive work relationships is not only the way to be more effective, its necessary for being a leader and getting ahead.
Current Focus
- Foundational AI model development and evaluation
- Applied AI systems and product delivery
- Principal-level platform architecture
- Cross-time-zone technical leadership
Education and Career Snapshot
2024 - Present
Tilt Finance
- Staff Software Engineer
2020 - 2024
Octopus Deploy
- Senior Software Engineer
2017 - 2020
Zendesk
- Data Science Engineer (AI Team)
2014 - 2017
University of Melbourne
- PhD, Developmental Genetics
2010 - 2013
University of Connecticut
- MSc, Genetics, Genomics, and Bioinformatics
2007 - 2010
University of Connecticut
- BSc, Molecular and Cell Biology