Brought to you by:
• Sentry – application monitoring software considered “not b
Brought to you by:
• Sentry – application monitoring software considered “not bad” by millions of developers sentry.io/pragmatic
• turbopuffer – a vector and full-text search engine built on object storage. It’s fast, cheap, and extremely scalable turbopuffer.com/pragmatic
YouTube www.youtube.com/watch?v=8xBJ...
Spotify open.spotify.com/episode/5c1n...
Apple podcasts.apple.com/us/podcast/w...
Brought to you by:
• Antithesis – verify your system’s correctness without human review or traditional integration tests – and avoid bugs or outages antithesis.com/pragmatic
Few people care more about performant code than @cmuratori.bsky.social. Casey is
Few people care more about performant code than @cmuratori.bsky.social. Casey is a down-to-earth programmer, game dev, performance nerd. A refreshing episode about why the basics (still) matter, and why learning to read Assembly remains useful for anyone who cares about performance:
(cont'd)
Full article on Reuters: www.reuters.com/investigatio...
I asked two months ba
Full article on Reuters: www.reuters.com/investigatio...
I asked two months back: why is Meta seemingly deliberately destroying its engineering organization? This is why. My analysis, back then: newsletter.pragmaticengineer.com/p/why-is-met...
Well this is damning and explains SO MUCH about why Meta is losing its best engi
Well this is damning and explains SO MUCH about why Meta is losing its best engineers; why morale is rock-bottom; and why an engineering culture built up in 20 years was destroyed overnight.
Zuck + execs wanted to lay off ~60% of staff.
When the business does better than ever.
Hard to believe but there was a time when smoking was promoted to be healthy / h
Hard to believe but there was a time when smoking was promoted to be healthy / harmless, even, pregnant women encouraged to smoke, 1920s-1950s. Took ~30 years (!!) to figure out how harmful it was.
How long will it take to figure out the impact of AI on health, for those using it all day?
@brettcodes.bsky.social appreciated watching your video on why you stopped using
@brettcodes.bsky.social appreciated watching your video on why you stopped using AI. Recently talked with Casey Muratori who also doesn't use it. Could I message you somewhere? Or you can also send me DM.
Brought to you by:
• Google Cloud Run – run untrusted agent code without the se
Brought to you by:
• Google Cloud Run – run untrusted agent code without the security anxiety. Cloud Run sandboxes deliver hyper-isolated, ephemeral execution environments that spin up in milliseconds. Check them out: docs.cloud.google.com/run/docs/cod...
Brought to you by:
• Antithesis – verify your system’s correctness without huma
Brought to you by:
• Antithesis – verify your system’s correctness without human review or traditional integration tests – and avoid bugs or outages. antithesis.com/pragmatic
• Sentry – application monitoring software considered “not bad” by millions of developers sentry.io/pragmatic
(cont'd)
If you’ve ever opened Chrome DevTools, or optimized a page for Core Web Vitals,
If you’ve ever opened Chrome DevTools, or optimized a page for Core Web Vitals, you’ve used software built by @addyosmani.bsky.social. We start from when he built a web browser from scratch, at 16, through his 14 years at Google, to what he sees working (and not working) building software with AI:
From Chrome DevTools to AI Engineering, with Addy Osmani
Addy Osmani spent more than 14 years at Google, working on Chrome, DevTools, Core Web Vitals, and most recently, AI developer experience.
If you've ever opened Chrome DevTools, or optimized a page for Core Web Vitals, you’ve used software built by Addy Osmani. In this episode, I sit down with Addy and we talk about his path from building a web browser aged just 16 to becoming a director at Google. We discuss what he learned from building tools for millions of developers, Google’s engineering culture, and why he continued doing hands-on coding work as a manager. We also get into how he works w
and thank you, glad you enjoyed it, I try to do more of them in-person as it flo
and thank you, glad you enjoyed it, I try to do more of them in-person as it flows better. Mario and Armin have a great vibe together (and I love how grounded both of them are)
I don't know, and there's not much of a way to be able to tell, beyond anecdotes
I don't know, and there's not much of a way to be able to tell, beyond anecdotes.
Anecdotally, when someone walks up to me on a conference, a year ago it was about 50/50 if they knew the newsletter or the podcast. This year in London and SF it was more 30/70 newsletter/podcast, interestingly!
This is such a cool list, published by Brex, based on startups' newsletter spend
This is such a cool list, published by Brex, based on startups' newsletter spending patterns.
The Pragmatic Engineer one of the most frequently expensed newsletters, and it was, indeed, massive growth within startups over the last year:
Thank you to everyone reading and supporting!
It's now the second blog post in a day sent over by an early-stage VC-startup I
It's now the second blog post in a day sent over by an early-stage VC-startup I tried to read, but could NOT, after a few sentences.
Verbose, weird... AI-generated.
These companies think they made "content" more efficient. No: it's lower quality, slop +I no longer care about the startup
Brought to you by:
• Buildkite — the CI platform trusted by OpenAI, Anthropic,
Brought to you by:
• Buildkite — the CI platform trusted by OpenAI, Anthropic, Cursor, Meta, Uber, NVIDIA, Airbnb and many more. buildkite.com/pragmatic
• WorkOS — make your app and agents Enterprise Ready, with SSO, SCIM, RBAC, and more. workos.com
• Spotify: open.spotify.com/episode/40QO...
• Apple: podcasts.apple.com/us/podcast/s...
Brought to you by:
• Antithesis— turbocharge testing of your systems by running your whole system under aggressive fault injection. antithesis.com/pragmatic
(cont'd)
In 2025, it was rational to be skeptical about whether AI would change the futur
In 2025, it was rational to be skeptical about whether AI would change the future of software development. In 2026, it's not, anymore:
We talked about this with @charity.wtf, one of my favorite voices in tech. We get into a LOT of things:
• YouTube: www.youtube.com/watch?v=HC8T...
(cont'd)
The analysis of how Meta's leadership seemingly deliberately started to destroy
The analysis of how Meta's leadership seemingly deliberately started to destroy their eng org: newsletter.pragmaticengineer.com/p/why-is-met...
Most damage can be fixed with enough money, but knowing you're a cog who leadership can fire or reassign any time won't go away. Mercenary culture?
What I'm hearing from ~3 such stories:
Meta went too far firing people. They ar
What I'm hearing from ~3 such stories:
Meta went too far firing people. They are now doing MASSIVE counter-offers like to those handing in resignations. Some accept it, most just use it for more $$ at New Job, desperate to get out.
It was such a predictable outcome btw:
Stop being skeptical about AI for development with Charity Majors
In 2025, it was rational to be skeptical about AI, but in 2026 it’s clear that AI is changing all of the industry, and there’s less and less place for skepticism. This take is from one of my favorite voices in software reliability and observability: Charity Majors, CTO and cofounder of Honeycomb, co-author of Observability Engineering. (Note: the second edition of Observability Engineering is out, and it’s pretty much a full rewrite of the book, I recommend grabbing it if you’re building reliable systems)
In this episode, I sat down with Charity to discuss how her thinking on AI has evolved,
The comment on LI. www.linkedin.com/feed/update/...
I just stop taking people s
The comment on LI. www.linkedin.com/feed/update/...
I just stop taking people seriously who prove they are either unable or unwilling to form their own thoughts. Sad and surprising to see people whose job used to be thinking+writing outsource thinking+writing to AI.
I’m used to AI slop on LinkedIn, but not to AI slop by folks with impressive pro
I’m used to AI slop on LinkedIn, but not to AI slop by folks with impressive professional background. Saw a fully AI-written post from an ex-VPE at Microsoft (15 years there) and decided to say what many think when they see AI nonsense, but usually just scroll by.
Bought a new Dyson fan, set it up connected to the Wifi, and then it suddenly tu
Bought a new Dyson fan, set it up connected to the Wifi, and then it suddenly turned off for minutes, unable to turn back on.
Turns out, it was spontaneously updating itself.
This is not how a household appliance should update, I’m sorry… why do we accept such sloppiness?
Interesting to see how any and all mid-sized and above companies who take the "A
Interesting to see how any and all mid-sized and above companies who take the "AI opportunity window" seriously have all built their own internal agent platform and one or more of agentic coding harness, agent sandboxing, model hosting, evals platform
Building AI infra > buying right now
If you work in tech, you'll want to read the ~40-page short novel, Profession, b
If you work in tech, you'll want to read the ~40-page short novel, Profession, by Isaac Asimov.
If the availability of intelligence was instant, and learning things the "hard way" was considered a waste: what would the things most valued by society be?
Hard to believe it's a 1957 novel.
Account still active 3 days later and is taking over accounts with the calendar
Account still active 3 days later and is taking over accounts with the calendar scam. Here’s someone almost a victim.
I couldn’t report it because there’s no X account impersonated. I tried!
x.com/sidi_jeddou_...
Google Workspace's ads for paying customers could not get any more annoying
I D
Google Workspace's ads for paying customers could not get any more annoying
I DO NOT USE AI AT ALL and I get upsells to "expand my AI access"
WTH. I want ZERO Gemini for my Workspaces suite, this is some bad joke. Even if you pay you get ads, just so maddening with Google
(If you have had good experiences with engineering leadership coaching with some
(If you have had good experiences with engineering leadership coaching with someone not listed on this page, but active and you can personally vouch for them, drop their details and I'll add them to this page.)
My page created in 2021 "coaches and mentors for engineering leaders" is seeing
My page created in 2021 "coaches and mentors for engineering leaders" is seeing both an increase in traffic, and folks listed here are pinging me about getting more business through this channel (as well!) blog.pragmaticengineer.com/coaches-and-...
Recent trend I'm hearing about: coaching demand for engineering leaders (done by
Recent trend I'm hearing about: coaching demand for engineering leaders (done by experienced eng leads offering paid coaching) is picking up after the last ~2 years being pretty quiet.
Clearly to do w rapid industry change, and lots of folks strategizing "how do I thrive?"
@simon-hill.bsky.social hey Simon - ai cannot DM you but wanted to let you know
@simon-hill.bsky.social hey Simon - ai cannot DM you but wanted to let you know someone is impersonating you on X. Tried to report the account but cannot do so because the impersonated account needs to be on that site.
Warned others on X: x.com/gergelyorosz...
I do think we will always have some level of fragmentation with specialized tool
I do think we will always have some level of fragmentation with specialized tools fwiw. And when a specialized tool works really really well no one minds, at least no one has minded in the past
While the idea is pretty obvious and recognizing that there's a bunch of fragmen
While the idea is pretty obvious and recognizing that there's a bunch of fragmentation in how work is done, it's hard to see if yet another product that has no context in any of these tools can "win"
Or if it will be an existing tool with lots of context also allowing to import context and "win"
Interesting to see so many apps / companies launching with the positioning to ga
Interesting to see so many apps / companies launching with the positioning to gather CONTEXT in a single space. Pitches like:
"work is scattered across many apps / websites / windows. What if..."
"you keep moving between tools / sites / docs to get your work done. What if..."
One of the easiest ways to stand out when writing or creating videos or posting:
One of the easiest ways to stand out when writing or creating videos or posting:
DONT. USE. AI.
Suddenly you sound distinctively different from 80%+ ppl who do and output the same verbose + nicely formatted stuff that feels all the same (+ most ppl stop paying attention midway)
There’s a popular theory that AI will finally make formal verification mainstream because mathematical proof of correctness will be needed when machines write most or all of the code. But will this happen? Today, I’m talking with one of the best people to tackle the prediction. Hillel Wayne is a formal methods consultant, educator, and author, who’s deeply interested in software history.
In this episode of Pragmatic Engineer podcast, I sit down with Hillel to compare software engineering with traditional engineering, discuss where formal methods fit into modern software development, and we e
Last month in @pragmaticengineer.com about how Meta is destroying its engineeri
Last month in @pragmaticengineer.com about how Meta is destroying its engineering culture: newsletter.pragmaticengineer.com/p/why-is-met...
Last week in Semi Analysis on how Meta's infra org needs a reset: newsletter.semianalysis.com/p/metas-infr...
Meta's performance review culture is becoming very problematic for its engineeri
Meta's performance review culture is becoming very problematic for its engineering culture.
I said this a month ago in @pragmaticengineer.com (screenshot #1), and Semi Analysis said the same last week (screenshot #2).
Engineers at Meta are world-class at optimizing for PSC...
From a friend who joined a fast-growth, later-stage startup:
"I was wondering w
From a friend who joined a fast-growth, later-stage startup:
"I was wondering why no one is doing PRDs here. Then I realized that the Head of Product and most PMs are 25-year-olds who don't have the attention span to *read* even a 1-page doc.
You lose them after 3 bullet points."
Eg eg Linear (originally: planning the work), Sentry (originally: logs), I assum
Eg eg Linear (originally: planning the work), Sentry (originally: logs), I assume soon incident management tools will do this etc
I wonder if we have a future where you can "one-shot" fixes from any and all of these tools: and you just do it in which you happen to be working at the given time?
And with AI tools, doing this is easy enough at any part of the stack. Eg for Co
And with AI tools, doing this is easy enough at any part of the stack. Eg for Codex you could use an MCP for your error service (Sentry or not) to do the same.
I also observe all tools part of the dev toolchain building tools to be able to push code in the end: (cont'd)
OK this made my life easier: my backend (for my admin portal) had an error poppi
OK this made my life easier: my backend (for my admin portal) had an error popping up, and Sentry was bugging me about it. Sentry now built an AI agent (called "Seer") that can analyze the root cause, make a plan to fix, draft a fix, then open a PR. Just pushed it to prod after review!
The benefits of longer tenure, using first principles to build durable software – and why founders should be cautious when raising VC money. turbopuffer CEO Simon Eskildsen at AI Engineer's World Fair, in San Francisco, in July 2026, interviewed by Gergely Orosz.
See the accompanying The Pragmatic Engineer deepdive: Pushing software engineering limits with “napkin math” at https://newsletter.pragmaticengineer.com/p/pushing-software-engineering-limits
Timestamps:
00:00 Intro
02:00 Competitive programming
05:18 Joining Shopify
08:28 Infra at Shopify
10:59 toxiproxy
15:30 The Napkin Math proje
These Google Workspace ads (for *paying* customers!) are driving me crazy
I don
These Google Workspace ads (for *paying* customers!) are driving me crazy
I don't know why I would need a "virtual receptionist", what "Voice Standard" is, who "my callers" are, and why I would need to "guide them automatically"
Is Google Workspace losing its mind? What next: proper ads?
Bt "Workplace" I am referencing "Meta's internal Facebook for work" that doesn't
Bt "Workplace" I am referencing "Meta's internal Facebook for work" that doesn't really have a name (Facebook for Work?) and based on which Workplace was built. But it's not the same.
bsky.app/profile/gerg...
I am referencing Facebook at Work, the product Facebook/Meta always had and base
I am referencing Facebook at Work, the product Facebook/Meta always had and based on which they launched Workplace (a different product). They shut down Workplace, but keep using their "internal Facebook" themselves.
Morale is slow low b/c Meta had been on a roll to destroy the engineering cultu
Morale is slow low b/c Meta had been on a roll to destroy the engineering culture built up over ~20 years with e.g. layoffs (when the business is booming), forced reassignments of devs to data labeling, tracking keystrokes etc. Covered more here: newsletter.pragmaticengineer.com/p/why-is-met...
So a bunch of devs started to call out, in the comments. Savage comments, just n
So a bunch of devs started to call out, in the comments. Savage comments, just no longer caring.
And the result is... these AI-written slop posts are... gone from Workplace!
** Workplace not really an internal LinkedIn, more like their own version of FB, used to communicate projects, ideas etc.
OK this is a HILARIOUS story I heard from inside Meta:
You know when Director+
OK this is a HILARIOUS story I heard from inside Meta:
You know when Director+ folks post AI slop as "thought leader" posts? But everyone is too polite to say anything. Happened a ton inside Meta's internal "LinkedIn" (Workplace**)
BUT morale is so low that lots of devs no longer gave a damn. So:
My kids keep pointing out how so many ads involving kids are AI-generated. This
My kids keep pointing out how so many ads involving kids are AI-generated. This is one example from a government agency. It’s in so many other places as well tho with kids’ posters
Kids that look more like adults, have 4 fingers…is this the “modern” advertising? All AI?
The current state of Spotify's publisher's page. Comments stopped loading, every
The current state of Spotify's publisher's page. Comments stopped loading, everything stopped loading, links point to dead pages... a truly fascinating case study of a product falling apart
AFAIK they value software engineers very much there
I exchanged emails with the
AFAIK they value software engineers very much there
I exchanged emails with their dev teams who assured me reliablity is top of mind
Then nothing happened, it was a waste of even emails
I don't think this is about valuing labor or not. Feels like reliablity not a concern
All of this comes after Spotify was bragging how they use AI for software develo
All of this comes after Spotify was bragging how they use AI for software development, publish a massive number of PRs to prod per day.
But reliablity has been going down for me. To the point that today this is what my portal looked:
The last 7 weeks, Spotify had 4 outages when publishing my podcast episodes (!!)
The last 7 weeks, Spotify had 4 outages when publishing my podcast episodes (!!). After outage #3, I asked the Podcasts team to share how they are improving reliablity. They didn't share - and it just got worse.
So I am no longer publishing video podcasts on Spotify, but redirect my audio RSS.
One thing I am hearing that is top of mind for many eng leaders
"What are we do
One thing I am hearing that is top of mind for many eng leaders
"What are we doing to deal with this ongoing increase in code review load?!"
Everyone is seeing it, no one seeing a solution, but lots of experiments from better tooling to starting to admit code review wont scale
Brought to you by:
• Buildkite — the CI orchestration platform built for reliab
Brought to you by:
• Buildkite — the CI orchestration platform built for reliable scale. Used by OpenAI, Anthropic, Cursor, Meta, Uber, Ramp, Nvidia, Airbnb and many more. buildkite.com/pragmatic
• Sentry — application monitoring software built by developers, for developers. sentry.io/pragmatic
Knowing how LLM contexts work and how to work around context limitations – aka “
Knowing how LLM contexts work and how to work around context limitations – aka “context engineering” – is becoming more important for software engineers working with LLMs. Dex Horthy coined the term "context engineering," and shares practical details on it:
YouTube: youtu.be/Usufn8IQJgw
(cont'd)
Knowing how LLM contexts work and how to work around context limitations – aka “context engineering” – is becoming more important for software engineers working with LLMs. Let’s look into what works and what doesn’t, today.
In this episode of The Pragmatic Engineer podcast, I sit down with the CEO and cofounder of HumanLayer, Dex Horthy, who coined the term “context engineering”. We discuss the ideas behind this context engineering, harness engineering, loop engineering, software factories, why his approach to AI-assisted software development has evolved, and how HumanLayer is helping enginee
The amount of AI writing on OpenAI's docs makes me sick.
Filler sentences for n
The amount of AI writing on OpenAI's docs makes me sick.
Filler sentences for nothing. Mannerisms and phrases humans would not write. Has a human even read this?
And all of this will change how other docs are written, and how we all talk-for the worse IMO.
🤮
developers.openai.com/cookbook/exa...
Yes a good one! Might steal it… I find it useful when I’m stuck (but it doesn’t
Yes a good one! Might steal it… I find it useful when I’m stuck (but it doesn’t happen too often so less reason to reach for it, plus I’m conscious of skills athropy)
I want to know the capability of tools available, in anything I do - perhaps for
I want to know the capability of tools available, in anything I do - perhaps for better, this is very very far off from usable
At the same time I now see / read so much text reeking of LLMs in my inbox, social media, online on blogs etc, and it is bad yes
No appearances yet. If you've seen Gergely Orosz as a guest somewhere, hit + suggest to add it.
