Writing @Pragmatic_Eng, the #1 technology newsletter on Substack. Author of @EngGuidebook. Formerly Uber & Skype.

Amsterdam, The Netherlands
Joined April 2009
Just wow: Brex looked at what paid newsletters corporate credit card customers at 30,000+ companies subscribe, and @Pragmatic_Eng is the 3rd most popular, and the most popular one for software engineering. Right after Stratechery and Lenny’s Newsletter:
I most certainly have 206 files or more that are visible externally - and they have nothing sensitive in them. That's why I share externally... At this pace, Google will turn into Adobe, deploying dark patterns at an increasing scale It's small lies like this that start it
1
33
I am so annoyed at scare tactics as a dark pattern trying to get you to upgrade your SaaS subscription. This is Google Docs doing so. I am already a paying customer - and no, I don't have "sensitive files" shared externally. Such a cheap way to try to get more $$, Google
11
94
Interesting to see how companies like Anthropic hire key people very differently to many other tech companies: This position pays $600-650K in base, plus stock. Most tech co's would engage with an exec recruiter firm to fill the role Anthropic's CPO just shares it on LinkedIn
Chris is such an inspiration. How can you not have deep respect for perpetual builders who, after all these years, still love this shit?
Chris Lattner (@clattner_llvm) is one of the most influential engineers of the past two decades. He created LLVM, Swift, contributed to TensorFlow, and created the Mojo programming language. What was the story about creating Swift - and why did he face resistance inside Apple when wanting to replace Objective C? What did he learn at Tesla, Google and CPU maker SiFive, that led him to working on Mojo at Modular? We cover these and many more in today's episode. Watch or listen: • YouTube: piped.video/watch?v=Fxp3131i… • Spotify: open.spotify.com/episode/2Nk… • Apple: podcasts.apple.com/us/podcas… Brought to you by: •⁠ @statsig ⁠ — ⁠ The unified platform for flags, analytics, experiments, and more. statsig.com/pragmatic •⁠ @linear – The system for modern product development. linear.app/pragmatic?utm_sou… My favorite quote from Chris in this episode: “I believe in the power of programmers. I believe in the human potential of people that want to create things. And that’s fundamentally why I love software is that you can create anything that you can imagine.”
2
18
“Fluff-writing is at all-time high” Agree It’s also what social media algorithms reward (plus outrage), sadly
Replying to @GergelyOrosz
saying this as a fellow product marketer in dev tools: benchmarks and fluff writing are at an all-time high. sadly it’s all coming from the top for pure optics. every optimization hack is just optics, ask them to share the archi with benchmarks or a video grab of the actuals, and there will be crickets.
4
1
67
Welcome to accounts that discovered that traffic on their social media account leads to business And how making things up —> more traffic And starting to optimise for this - who can check anyway? Yes ~3 stories per day of how “we cut back our cloud spending by $$$”
Replying to @GergelyOrosz
I had the same issue with this account. I thought he is actually sharing something and commented once. Now he is in my feed everyday with same garbage madeup stories
2
3
1
87
After the made-up story of how they saved their startup by spending $1,800 per month on AWS the next made-up story is suddenly spending $52/month on AWS Sure
Cut our AWS bill from $52K to $18K per month. Took 3 weeks of detective work. The audit: - Started with AWS Cost Explorer - Noticed NAT Gateway was $8K/month - Data transfer was $12K/month - RDS storage was $6K/month What we found: - Logs were being sent to S3 via NAT Gateway - Should have used VPC Endpoint (free) - RDS had 14TB of automated backups - Retention was set to 90 days - Old snapshots from deleted databases The fixes: - VPC Endpoints for S3 and DynamoDB - Backup retention to 7 days - Deleted 200+ orphaned snapshots - Moved development to Spot instances
2
76
The amount of people believing made up tweets like this below is concerning For some reason no one stopped to check what “our startup” means on a consulting account In the next tweet it’s “We were paying $15K/month for an RDS Postgres instance.” Blue check != it’s real
Kubernetes migration almost killed our startup. Where we were: - 8 EC2 instances - Ansible for deploys - Boring but working - $1200/month AWS bill Why we migrated: - New investor wanted 'cloud-native' - Engineers wanted K8s experience - Competitors were using it - Seemed like the future 6 months later: - 3 engineers spending full-time on K8s - AWS bill at $4500/month - Deploys took longer than before - More outages, not fewer - Product development stalled We rolled back: - Moved to ECS Fargate - 2 week migration - Back to $1800/month - Engineers back on features K8s is amazing for scale. We weren't at scale. Technology should solve problems you actually have.
You understand this when you see it! Not all good and fun businesses are shiny and VC-funded Also this is not for everyone, but it can be satisfying if you enjoy working with people + staying on the cutting edge and bringing people with you Niko sees it:
Replying to @GergelyOrosz
Yes, I was invited to do just this for a local 50-dev company, and I am still supporting them with monthly checkups, as these things are just hard for those outside the AI echo chamber.
2
2
11
Source of my information on demand existing: Talked to a few Director+ folks at “traditional” companies (think: 50+ year old, ones like banks, airlines etc with massive tech teams) They are pouring money into AI pilots, licenses, and want to find training They gave up on training (for now) as there is none
2
2
14
And before you roll your eyes at “training” Done well, workshops that get people to try new tools, new ways of working, and having “aha!” moments are awesome It’s hard to do well, and needs to be continuously updated as tools change: so it’s hard to do it right! Again, demand is there, as are budgets…
2
16
A free, non-VC funded lifestyle business idea: AI training / workshops for devs Massive demand from more traditional companies that have budget for this and are seeing devs not really picking up these tools. They want to invest Doesn’t exist as far as they are aware
18
5
2
108
Gergely Orosz retweeted
Wonderful talking with Gergely about building large scale systems, and how my work at 🍎 building LLVM/Swift for CPUs connect to my journey to unify AI compute with MLIR + Mojo🔥. Clarification: Apple's approach is nearly unique, being willing to take very hard swings at very big problems. Asking hard question is part of the process! Apple leadership asked many hard q's in Swift development, but they were the right questions to ask, and helped shape Swift into the success that it became. I credit leadership with asking hard questions, setting a high bar, and demanding excellence. This approach is what allowed it to bet full force on such a radical shift in its developer ecosystem - this required conviction, alignment, and determination at all levels. This is made possible through a very deliberate set of leadership principles. It also helped that Swift is a good thing of course :-) I cherish my time and learnings from Apple, and consider myself to carry "Apple DNA". It's a magical team!
Chris Lattner (@clattner_llvm) is one of the most influential engineers of the past two decades. He created LLVM, Swift, contributed to TensorFlow, and created the Mojo programming language. What was the story about creating Swift - and why did he face resistance inside Apple when wanting to replace Objective C? What did he learn at Tesla, Google and CPU maker SiFive, that led him to working on Mojo at Modular? We cover these and many more in today's episode. Watch or listen: • YouTube: piped.video/watch?v=Fxp3131i… • Spotify: open.spotify.com/episode/2Nk… • Apple: podcasts.apple.com/us/podcas… Brought to you by: •⁠ @statsig ⁠ — ⁠ The unified platform for flags, analytics, experiments, and more. statsig.com/pragmatic •⁠ @linear – The system for modern product development. linear.app/pragmatic?utm_sou… My favorite quote from Chris in this episode: “I believe in the power of programmers. I believe in the human potential of people that want to create things. And that’s fundamentally why I love software is that you can create anything that you can imagine.”
7
22
351
"[AI coding tools like Cursor] increase my enjoyment, as a programmer" THIS is something we don't talk about enough. When a tool makes the job better - eg because it removes the toil that's a drag to do (e.g. refactoring, repetitive stuff) then it's already a big win!
Enjoyed this interview with @clattner_llvm Software engineering isn't going anywhere, and we need more engineers wanting to become masters of their craft. AI coding tools can increase your enjoyment by removing the drudge work. They don't replace your thinking.
Gergely Orosz retweeted
Incredible interview with a fascinating engineer. I had no idea one programmer had a hand in so many projects I’ve come to appreciate.
Chris Lattner (@clattner_llvm) is one of the most influential engineers of the past two decades. He created LLVM, Swift, contributed to TensorFlow, and created the Mojo programming language. What was the story about creating Swift - and why did he face resistance inside Apple when wanting to replace Objective C? What did he learn at Tesla, Google and CPU maker SiFive, that led him to working on Mojo at Modular? We cover these and many more in today's episode. Watch or listen: • YouTube: piped.video/watch?v=Fxp3131i… • Spotify: open.spotify.com/episode/2Nk… • Apple: podcasts.apple.com/us/podcas… Brought to you by: •⁠ @statsig ⁠ — ⁠ The unified platform for flags, analytics, experiments, and more. statsig.com/pragmatic •⁠ @linear – The system for modern product development. linear.app/pragmatic?utm_sou… My favorite quote from Chris in this episode: “I believe in the power of programmers. I believe in the human potential of people that want to create things. And that’s fundamentally why I love software is that you can create anything that you can imagine.”
2
27
Apple is paying $1B to Google to use a whitelabelled Gemini to power Siri While Snap gets $400M from Perplexity to allow them to build AI search for Snap in a branded way Comes to show how a large platform giving visibility / distribution to another company is worth $$$
I like this move by $SNAP. Partnering with Perplexity and giving them distribution in return for $400M a year. The wording also suggest there might be more of these kind of deals coming.. After this aftermarket move $SNAP is now one of my biggest positions.
6
3
3
122
So many other Big Tech companies would never do this. They would be too worried about giving credit to engineers, and their PR and Comms people would be terrified “what if they spill some key details to competition?” So they remain closed off, and you and me will never learn about the interesting work they do - and the best of the best will have no desire to work there Their loss!
3
2
24
Anthropic is one of the few companies that “get” how people stop caring about corporate PR but are interested in in-depth stuff. It’s why I was able to share the insider story of how exactly Claude Code was built (talking with the founding engineers:) newsletter.pragmaticengineer…
Think of these roles as being a Substacker within Anthropic with access to real time data, insider knowledge of what’s happening at the frontier of AI + independence to write about the topics that matter most to you. These roles are for big thinkers and domain experts, not content marketers.
Chris Lattner (@clattner_llvm) is one of the most influential engineers of the past two decades. He created LLVM, Swift, contributed to TensorFlow, and created the Mojo programming language. What was the story about creating Swift - and why did he face resistance inside Apple when wanting to replace Objective C? What did he learn at Tesla, Google and CPU maker SiFive, that led him to working on Mojo at Modular? We cover these and many more in today's episode. Watch or listen: • YouTube: piped.video/watch?v=Fxp3131i… • Spotify: open.spotify.com/episode/2Nk… • Apple: podcasts.apple.com/us/podcas… Brought to you by: •⁠ @statsig ⁠ — ⁠ The unified platform for flags, analytics, experiments, and more. statsig.com/pragmatic •⁠ @linear – The system for modern product development. linear.app/pragmatic?utm_sou… My favorite quote from Chris in this episode: “I believe in the power of programmers. I believe in the human potential of people that want to create things. And that’s fundamentally why I love software is that you can create anything that you can imagine.”