Issue #742

Essential Reading For Engineering Leaders

August 14th, 2026

Friday 14th August’s issue is presented by Gauntlet

Online + on-site in Austin, TX · Cohort 7 starts September 14 · Fully funded

Most programs ask you to bet on them. You pay, you hope it pays off. Gauntlet runs the other way. It's a ten-week, fully funded sprint built to turn people who use AI into people who build with it as second nature, and it puts real resources behind that before you've paid a cent. Three weeks remote to ramp, then seven weeks on-site in Austin. Flights, housing, food, compute, and model access are all covered. No tuition. You bring the work, Gauntlet covers the rest, and the builders who come out the other side are seeing offers start at $200K+.

It isn't for everyone, and that's the point. If you're ready to stop experimenting and start shipping under real pressure, Cohort 7 is the one to aim at.

What you get:

  • Fully funded, zero tuition: flights, housing, food, compute, and model access all covered.

  • 3 weeks remote, then 7 weeks on-site in Austin building next to serious people.

  • Real projects on real deadlines, until building with AI is reflex instead of experiment.

  • Offers coming out the other side start at $200K+.

— James Stanier

tl;dr: “If starting a project is like takeoff, finishing it is like landing, demanding a completely different kind of skill and attention, and it’s where most of the risk accrues. After all, the hardest part of flying a plane is landing it. This article is about why endings are so hard, what goes wrong when you don’t manage them deliberately, and how to land your projects well.”

Management Productivity

tl;dr: “The cardinal rule of meetings with your direct reports is that you are always on the same side of the table. From the moment you walk in to the moment you leave, every single thing that your team does is a reflection and extension of you.”

Leadership Management

tl;dr: Online + on-site in Austin, TX · Cohort 7 starts September 14 · Fully funded. Most programs ask you to bet on them. You pay, you hope it pays off. Gauntlet runs the other way. It's a ten-week, fully funded sprint built to turn people who use AI into people who build with it as second nature, and it puts real resources behind that before you've paid a cent. Three weeks remote to ramp, then seven weeks on-site in Austin. Flights, housing, food, compute, and model access are all covered. No tuition. You bring the work, Gauntlet covers the rest, and the builders who come out the other side are seeing offers start at $200K+. It isn't for everyone, and that's the point. If you're ready to stop experimenting and start shipping under real pressure, Cohort 7 is the one to aim at. What you get: (1) Fully funded, zero tuition: flights, housing, food, compute, and model access all covered. (2) 3 weeks remote, then 7 weeks on-site in Austin building next to serious people. (3) Real projects on real deadlines, until building with AI is reflex instead of experiment. (4) Offers coming out the other side start at $200K+.

Promoted by Gauntlet

Event AI

— Steve Huynh

tl;dr: “There is a whole class of advice that gets repeated so often it stops registering with us. Somewhere along the way, we decide that anything this common must be for beginners, so we’re allowed to skip it. We nod, we say “Yeah, yeah, yeah,” and then we go looking for the real advice. You can spot this sort of advice by noticing (a) how fast you agree with it and (b) how rarely you do it.”

CareerGrowth

"Educating the mind without educating the heart is no education at all."

— Aristotle

— Fred Hebert

tl;dr: “The adoption of LLMs in software development has led countless organizations to rapidly change their practices and structures. Old methods are questioned, replaced, and repurposed as the economics around creating new code get shaken up. Because humans and LLMs aren’t interchangeable, the dynamics in play are also very different. Systems are systems, and so regardless of what is changing, there are known patterns on which we can draw to provide some guidance and warnings.”

Architecture AI

tl;dr: Brex went from duplicating 800+ microservices per preview environment to running only the services under test, saving about $2 million a year. Deploy times dropped from ~30-60 minutes down to less than five, and developer CSAT came in 28 points higher. Read the case study to see how their platform team scaled testing across hundreds of engineers and coding agents while slashing costs.

Promoted by Signadot

Tools+Setup CaseStudy

— Thariq Shihipar

tl;dr: “Workflows allow you to dynamically create harnesses that enable Claude to solve all of those problems and more natively inside of Claude Code. You can also share and re-use these workflows with others. In this article, I’ll cover my initial workflows experiences and learnings so you can take full advantage.”

— Gergely Orosz

tl;dr: “Building better AI models getting more focus than lower latency, owning the full stack from applications to building custom hardware, very different incentives to most tech companies in play, and more.”

— Vicki Boykis

tl;dr: “We should strive for simplicity. If we can’t, it could be because we don’t understand the problem ourselves yet enough to compress it. That’s cool. Understanding things is hard. Naming things is hard, for the same reason. But it should be a little harder. Like 10% harder. We should be more tired than the model, or at least a little less verbose.”

AI Communication

Managing Up

Hand-drawn by Manu. View the Null Pointer series.

Computer: Give your agent a computer.

Corsair: Unified integration layer for agents.

Diagram Design: Editorial diagram types for Claude Code.

Needle: Foundation model for tiny devices.

Sourcebot: Self-hosted tool to helps understand your codebase.

How did you like this issue of Pointer?

1 = Didn't enjoy it all // 5 = Really enjoyed it
1  |  2  |  3  |  4  |  5

Login or Subscribe to participate in polls.