Most engineering leaders are just developers pretending to know management. We all needed a playbook. So I wrote one ➡️ https://t.co/XHSAANW2Aa Readers say it changed the game. See for yourself.

That’s what I’m writing about in this week’s Global Talent Sprint: If you want it in your inbox before release day, you can sign up here:

You’re not a micromanager. You’re just filling the gaps. So you jump in on a code review to speed things up. You rewrite vague tickets. You fix blockers yourself. You jump into every code review “just to help.” It feels productive. But it’s building dependency. Because every

It’s now called “bad vibes coding.” Why? Non-technical people keep vibe coding features and handing them to engineers to “figure out.” It's the stuff of horrors for most software engineers. Don’t be surprised when your engineers don’t like your vibe.

That means calling a virtual war room. Getting everyone together to talk through things, solve hard problems, and figure shit out. It still works the same. We just have to be more deliberate about it. The only real difference? You can’t order pizza. 🍕

Remote doesn’t work because you can’t lock people in a room to figure shit out. When you’re in an office, it’s easy to throw everyone in a conference room, order pizza, and say: “We’re not leaving until we figure this out.” That urgency is harder to create remotely. It’s easier

The real challenge isn’t connecting AI to your data. It’s knowing where the line should be. ⚡Restrict access through an API layer, not direct SQL ⚡Validate intent, limit permissions, and log every query ⚡Treat read-only as the default, not the exception AI should assist you,

We keep talking about autonomous AI agents like they’re senior engineers. They’re not. They’re brilliant interns with no context and no fear of consequences. If you wouldn’t give a new hire production access on day one, why would you give it to an AI model?

AI and SQL don’t mix without a chaperone. Unless you enjoy turning your infrastructure into a dumpster fire. AI doesn’t understand performance. It doesn’t know the difference between a good query and one that locks your tables for ten minutes and sends your cloud bill through

Most offshore teams fail because they’re built like vendors, not teammates. When devs are embedded in your workflow, not managed from the outside, they perform like locals. I break it down in my new piece: How to Build a Global Team That Feels Local 👉 Read it here:

👉 Want it in your inbox before release day? Sign up here:

If your rituals only work when everyone is in the same room, they aren’t culture. They’re convenient. I’ve seen remote and hybrid teams with stronger culture than teams that share four walls. Because they built it on purpose—through structure, clarity, and connection. If you

If you treat developers like interchangeable resources, you’ll always be stuck in onboarding mode. When you treat them like long-term teammates, context compounds, trust builds, and velocity finally stabilizes. That’s the difference between renting output and earning outcomes.

AI can teach you how to fill a propane tank. It won’t give you the experience to not smoke while doing it. Knowing how to do something isn’t the same as knowing what not to do. Experience still matters. Judgment still matters. AI can teach the steps. It can’t teach the scars.

AI was probably invented by a dev who got tired of writing regular expressions.

There’s a big difference between a team that **ships** and a team that **scales**. A team that ships can push code. A team that scales can do it *consistently*, sprint after sprint, without burning out or starting over every quarter. I’m breaking down how to build that kind of

On a podcast yesterday, someone asked me: “How much context do software engineers really need?” My answer was simple. Usually, it only takes five extra minutes. Five minutes to write it in a ticket. Five minutes to explain it in a planning meeting. Five minutes to share why it

Because the best products aren’t built by handing things off. They’re built by planning together.

The truth is, product and engineering can’t be led separately. You can’t make product decisions without understanding how the system works. And you can’t make engineering decisions without knowing what actually matters to customers. That’s why the best product leaders have an

That’s when they have to ask: How will this scale? What does it break? Who needs to be involved? How much risk are we really taking on? When product skips that conversation, engineers feel like they’re being set up to fail. And honestly, they’re right. That’s why vibe coding

Product thinks a product plan is a real plan. Engineering knows it isn’t. A product plan says what to build. An engineering plan figures out how to make it real. That “how” is where all the hard stuff lives. The trade-offs, dependencies, risks, and decisions that make or break