Explore Johnny
LinkedIn ↗
Johnny
Startup Engineering Leader | Founding Engineer | Product Systems & AI-native Delivery
Startup engineering leader and founding-engineer type with 20+ years of experience turning ambitious ideas, messy operational reality, and business complexity into scalable software systems. I work at the intersection of product, engineering, delivery, and operations: shaping what should be built, why it matters, how it should work in practice, and how to get it delivered in a way that is fast, safe, and verifiable. I am most effective in startup and scale-up environments, where speed matters but so do trust, clarity, and long-term thinking. My background spans product shaping, architecture, operational workflows, integrations, reliability, and technical leadership. More recently, I have focused on AI-native software delivery: agent-friendly systems, explicit workflows, strong validation, and dark-factory approaches that improve execution without compromising quality.
Sunderland, UK
Selected work
View all work →In-house Engineering Transition
Established engineering standards, CI/CD discipline, and operational readiness.
Rails Marketplace Scale-Up
Grew monthly GMV from roughly GBP thousands to GBP millions while maintaining reliability.
Business-Critical Integrations
Improved delivery throughput with explicit appetites and risk containment.
Experience
View full experience →Explore My Profile
Founder / Founding Engineer
Building a product that turns static professional profiles into richer, more inspectable experiences for both people and AI agents.
Materials Market
First Engineer -> Head of Engineering (hands-on) / Tech Lead, Ruby on Rails
Joined as the first technical hire and helped transition the business from agency dependence to an in-house engineering function.
Kudocs
Tech Lead, Ruby on Rails
Led development of a LegalTech SaaS platform, owning delivery across feature work and operational maintenance.
Writing
View all writing →When the Agent Implements, the Human Stops Learning
This post argues that implementation is where developers learn which assumptions fail, where the domain is messier than expected, and how the solution should change. It makes...
Early Error Handling Gives Away an Agent Built the Feature in the Wrong Order
This post argues that speculative error handling is a review smell that often reveals a coding agent built too much before learning from a working path. It makes the case for...
The Capability Is Real. The Discipline Is Sold Separately.
This post argues that autonomous-agent demos prove capability, not production readiness. It makes the case that token limits, realistic repositories, verification, operational...
Published with Explore.