TL;DR
Get business pricing on networking and server gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Peter Mattis, co-founder and CTO of Cockroach Labs and an original creator of GIMP, appeared on The Pragmatic Engineer podcast to discuss distributed databases, his work on Gmail storage and Google’s Colossus, and how AI coding tools have made him more productive than ever. He argued AI can improve code quality and multiply the impact of domain experts.
Peter Mattis, co-founder and CTO of Cockroach Labs and one of the original creators of GIMP, said in a newly published interview on The Pragmatic Engineer podcast that he is more productive writing code today with AI tools than he was earlier in his career, when he regularly shipped roughly 100K lines of database-grade code to production each year before AI assistance existed. In the episode, Mattis traces his path from open-source software to Google — where he worked on Gmail and distributed storage — to founding a database company, and discusses how distributed systems stay fast, reliable, and correct at scale.
Mattis said AI has brought him back to hands-on coding after his role at Cockroach Labs shifted toward management. He stated he now feels even more productive than before, with no drop in quality, and argued that AI can improve quality and multiply the impact of domain experts. The episode also touched on the future of code review and how engineers can continue leveling up their skills in an era of AI-assisted development.
Several technical details from his Google years stood out in the conversation. According to Mattis, B-trees were used in the first version of Gmail: each incoming message was matched to a thread using the search index, and threads and their unread counts were tracked by B-trees. Gmail launched on April 1, 2004, offering 1GB of free email storage — an offer so generous at the time that many assumed it was an April Fool’s joke.
At Google, Mattis also worked on Colossus, the successor to the Google File System (GFS). He and the team pioneered Reed–Solomon erasure coding in a distributed file system. According to the interview summary, this approach stored data twice instead of three full copies as GFS did, reducing storage overhead by 33% while increasing redundancy. A recurring theme across the episode, Mattis noted, is that much of distributed databases and storage systems — from Gmail’s backend to CockroachDB’s range index — ultimately resembles a B-tree, a phenomenon documented in the paper The Ubiquitous B-Tree.
Why a Veteran CTO’s AI Productivity Claim Matters
Claims from experienced engineers like Mattis carry weight in the ongoing debate over whether AI coding tools help or hurt software quality. Unlike many anecdotal accounts from early-career developers, Mattis’s perspective comes from someone who shipped large volumes of production database code for years without AI assistance — making his reported productivity and quality gains a notable data point. His view that AI multiplies the impact of domain experts suggests experienced practitioners, not just beginners, may benefit most, which has implications for how engineering teams structure work and mentorship.
The episode also offers practitioners concrete engineering lessons: the tradeoffs in building distributed databases, latency fundamentals — Mattis keeps “speed of light numbers” in his head, noting that intra-zone network round trips dropped from milliseconds to roughly 100 microseconds since Colossus was built — and his observation that the fastest global packet route may be through space via Starlink laser links, since light travels faster in a vacuum than in fiber.
Top picks for "distribut databas peter"
As an affiliate, we earn on qualifying purchases.
From GIMP to Google to CockroachDB
Mattis created GIMP, the free and open-source image editor, with his college roommate Spencer Kimball; the very first version of the Google logo was made in GIMP. In 2001, Sergey Brin made Mattis a job offer after an interview, but Mattis initially declined because he lived in San Francisco and did not want the hour-long commute to Mountain View, joining another startup instead. When Google reached out again, he accepted, later working on Gmail and distributed storage.
Mattis recounted nearly abandoning GIMP before launch after seeing an announcement for a rival photo editor promising more features — a project that was never heard from again. His advice to founders: “There’s always going to be someone else working on your idea… assume that dozens of people have the same idea you have, but most won’t ship it.” He also described twice outperforming standard library data structures at Google: replacing std::map‘s red-black tree with a faster, more compact B-tree, and later building a Swiss Table hash map for Go that was faster than the built-in map and eventually merged into the Go standard library with help from the Go team.
“There’s always going to be someone else working on your idea. You can’t get dissuaded if they pre-announce it. To founders: assume that dozens of people have the same idea you have, but most won’t ship it!”
— Peter Mattis, co-founder and CTO of Cockroach Labs
What the Episode Does Not Substantiate
Mattis’s productivity and quality claims are personal impressions shared in an interview, not measured benchmarks; no specific metrics, project artifacts, or comparative data were provided to verify them. The technical figures cited — such as the 33% storage overhead reduction from Colossus’s erasure coding and the ~100 microsecond intra-zone round-trip time — are presented as interview recollections rather than published measurements, and the full details are only available in the episode itself. It is also unclear how broadly his experience with AI coding tools generalizes to other engineers or organizations.
Where to Follow the Conversation
The full episode is available on YouTube, Apple Podcasts, and Spotify, with a transcript and timestamps published on The Pragmatic Engineer website. Readers interested in the low-latency routing discussion Mattis referenced can seek out the paper ‘Delay is not an option: low latency routing in space’. For those tracking Mattis’s ongoing work, Cockroach Labs’ releases and engineering blog remain the primary sources for developments on CockroachDB.
Key Questions
Who is Peter Mattis?
He is co-founder and CTO of Cockroach Labs, the company behind CockroachDB, and an original creator of the open-source image editor GIMP. He previously worked on Gmail and distributed storage at Google.
What did Mattis say about AI and productivity?
He said AI tools have brought him back to writing code after shifting toward management, and that he feels more productive than when he shipped roughly 100K lines of database-grade code annually pre-AI, with no drop in quality. He believes AI can improve quality and multiply the impact of domain experts.
What role did B-trees play in early Gmail?
According to Mattis, the first version of Gmail used B-trees in its storage layer: incoming messages were matched to threads via the search index, and threads and unread counts were tracked using B-trees.
What was the Colossus improvement he described?
Colossus, GFS’s successor, used Reed–Solomon erasure coding pioneered by Mattis and his team. Data was stored twice instead of in three full copies, which reportedly cut storage overhead by 33% while increasing redundancy.
Where can I listen to the full interview?
The episode is available on YouTube, Apple Podcasts, and Spotify, with a transcript and timestamps on The Pragmatic Engineer newsletter website.
Source: rss
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
