Skip to main content

Steve Kinney

Steve Kinney

I’m a software engineer, educator, and engineering leader based in Denver. Most recently, I spent a year as an Entrepreneur in Residence at Redpoint, where I built AI-native developer tools: Weft, an open-source durable execution platform; Tribunal, a platform for managing code review agents; Cinder, a component library for quickly prototyping AI interfaces; and an agent runtime embedded in a Chrome extension.

Before that, I joined Temporal as employee #35 and its founding front-end engineer. I architected the open source Temporal UI, the Temporal Cloud product experience, the self-service signup flow, a design system spanning product and marketing, and a Visual Studio Code extension. Temporal grew from a Series A startup into a company valued at $5 billion with more than 1,000 cloud customers while I was there, so I got to work across the full arc from “no interface exists yet” to enterprise customers depending on it in production.

My role at Temporal was Head of Engineering for Frontend and Developer Tools. I built and led the team from just me to eight engineers, stayed technical through architecture and design reviews, and made the call to hand major product surfaces to staff engineers instead of becoming the bottleneck. Along the way, I promoted four engineers and mentored my own successor into the role.

Earlier in my career, I was the very first Front-End Architect™ at Twilio and SendGrid. I also led the team that shipped the new Twilio SendGrid Marketing Campaigns product in under six months after the initial estimates were hovering around two years. Last I checked, that product was still in production and bringing in roughly $15 million in annual revenue.

I still care deeply about front-end architecture, but my work has never really been just front end. The common thread through all of it is making complicated systems usable: distributed systems, developer platforms, and now AI systems. I tend to live at the intersection of product thinking, technical architecture, and developer experience.

That also shows up in my writing. Lately, I’ve been researching and publishing on agent loops, memory systems for AI agents, AI gateways and durable workflow systems, prompt engineering across the OpenAI, Anthropic, and Gemini APIs in this post, MCP Apps and AI tooling, and what it takes to run untrusted code safely.

Today, I’m an instructor at Frontend Masters, where I teach courses including Open Source AI with Python & Hugging Face and Cursor & Claude Code: Professional AI Setup. I also founded the Front-End Engineering program at Turing School.

I actually started my career in public education. I was an NYC public school teacher in high-need schools across Manhattan, Brooklyn, and Queens. I technically had my principal’s license, but I never ended up using it.

I’ve co-founded DinosaurJS, a JavaScript conference in Denver, spoken at events like JSConf and RailsConf, and authored ”Electron in Action“—though technology moves faster than books, so consider that a historical artifact at this point.

Recordings

I am lucky enough to teach a bunch of courses with my friends at Master.dev. We've been working together since 2016. Before I was a teacher, I was a customer back when I was learning the ropes. I can't recommend them highly enough. You can find the most up-to-date list here.