Who Is Mat Ryer?
Before most of the internet knew what Go was, Mat Ryer was already building with it.
The year was somewhere around 2011, Go's v1 release was still months away,
and the language still had os.Error instead of the interface we know today.
Mat didn't mind. He had found something elegant - a language that was simple enough to hold in your head
and fast enough to mean it. He has been writing it ever since.
Today he is Engineering Director at Grafana Labs, building AI agents that help developers make sense of their systems at scale. It is a long way from the Commodore Amiga his father brought home when Mat was six years old - and also, somehow, a very short distance. The obsession is the same. What changed is the scale.
Mat grew up in rural Nottinghamshire, England, the kind of place where you make your own entertainment. His entertainment was a ZX Spectrum, then an Amiga, then every version of every computer that came after. By the time he was 18 he was already working at a local agency in Mansfield, building websites before most people had one. He moved to London in 2006 - his wife Laurie had accepted a job at the Science Museum - and the city's tech scene never let him leave.
At BT he found something he hadn't expected: the value of working with other people who were genuinely good. The agile practices he picked up there, in a room full of talented engineers and managers, still shape how he thinks about software today. It is not the tool. It is the discipline of thinking clearly, working iteratively, and not making things harder than they need to be.
"I've been writing HTTP services in Go for more than 13 years - and I still have new things to say about it."- Mat Ryer, Grafana Labs Blog, 2024
That philosophy runs through everything Mat touches. His most famous tool, xbar (originally called BitBar), is a macOS menu bar utility that puts the output of any script directly into your toolbar. The idea is almost absurdly simple. The execution is flawless. It has more than 18,000 GitHub stars. The original was written in Objective-C; Mat eventually rewrote it in Go, because of course he did.
His testing tools tell the same story. moq generates Go interface mocks using
go generate - clean, correct, no drama. The is framework is described
on its own GitHub page as "Professional lightweight testing mini-framework for Go."
That word choice - professional, lightweight - is pure Mat Ryer. He does not like ceremony.
He does not like boilerplate. He especially does not like code that requires explanation
when the code itself could just be clear.
His book, Go Programming Blueprints, published in 2015 and updated in a second edition, became a standard recommendation for developers learning how to build real things with Go. Not toy examples. Not contrived exercises. Real-world patterns for real-world problems. The same year, he co-wrote Go: Building Web Applications with Nathan Kozyra. Between them, the two books covered a lot of the practical territory that the official documentation left for readers to discover on their own.
340 Episodes. 8 Years. One Community.
The Go Time podcast ran for 340 episodes over eight years, ending in December 2024. For much of that run, Mat Ryer was its most consistent presence - not just as a host, but as the person who made the format feel like a conversation between friends rather than a roundtable of experts trying to sound impressive.
The format varied. Sometimes it was deep technical dives - stack internals, HTTP service patterns, the future of the Testify library. Sometimes it was confessional: "The se7en deadly sins of Go," an episode that became an instant community favorite. Sometimes it was genuinely strange: Mat playing piano mid-episode on Changelog & Friends, or hosting a live game show called "Gophers Say!" on stage at GopherCon EU Berlin in 2024, turning a technical conference into something that looked suspiciously like a game show.
That range - from rigorous to ridiculous - is the thing that made him irreplaceable in the community. Go has always been a language that takes clarity seriously. Mat Ryer is a person who takes clarity seriously, and also knows that the best way to get people to listen is to give them a reason to laugh.
After Go Time ended, he moved to Changelog & Friends, where his appearances have continued in 2025 and 2026 with the same mix of technical insight and structured chaos. The format changes. The character does not.
Go Time Greatest Hits
- Ep. 322 - "How Mat writes HTTP services in Go" (13 years of practice, distilled)
- Ep. 294 - "The se7en deadly sins of Go"
- Ep. 288 - "A deep dive into Go's stack"
- Ep. 340 - The Series Finale (Dec 2024)
- Go Time #139 - The future of Testify
- Live: "Gophers Say!" at GopherCon EU 2024
Recent Appearances
- #123 (Jan 2026) - 2026 predictions & the GitHub problem
- #107 (Aug 2025) - Git tooling & musical interludes
- #90 (Apr 2025) - "Hello, Matworld!"
- #75 (Jan 2025) - Piano & open source recognition
Building the Future of AI Observability
At Grafana Labs, Mat's current work sits at the intersection of Go, AI, and observability. He leads engineering on Grafana's AI agent initiatives - tools that help developers navigate complex cloud systems without needing to already know what they're looking for. In March 2025, he teased publicly that Grafana had "something new cooking that may just change everything." The details arrived in the form of Grafana Assistant, a context-aware LLM agent built directly into Grafana Cloud.
It is a natural evolution. Observability has always been about making invisible systems visible. AI is, in some sense, the same problem: making complex outputs understandable to the people who need to act on them. Mat has written about this on the Grafana blog - "AI for Observability," "AI in observability at Grafana Labs" - bridging the gap between what the systems know and what engineers can understand in the time they have.
His 2024 post "How I write HTTP services in Go after 13 years" became one of the most widely shared Go articles of the year. It is a masterclass in the value of lived experience: not just knowing the patterns, but knowing why the patterns evolved, which ones still hold, and which ones he has quietly abandoned. It reads like a conversation, not a tutorial.
Tools Developers Actually Use
The Long Game
os.Error was still a thing. Falls in love with the simplicity. Does not stop.The Scorecard
Moments Worth Knowing
os.Error was still in the language - before the API was stable,
before the toolchain was polished, before anyone was sure Go would catch on. He fell in love with its
simplicity anyway. By the time the rest of the industry discovered Go, he had already been writing it for years.