Who's the target?

Welcome to this installment of Veo, where every week, I share a bit about what myself and others are seeing on the cutting edge of software product development.

In this issue…

Did someone forward this to you? If you like it, you can sign up here.

Want to grab time on my calendar? You can do that here.

-John Cox

I really like to throw darts. I mean really, really like it.

I’ve had a dart board in my residence since I was very young - I can’t really remember not having one in dorm rooms, apartments, houses or offices. There’s something about the promise and elusiveness of perfection that keeps me coming back for more.

Naturally, throwing darts gets me thinking about targets. And that gets me thinking about a question I hear a lot when it comes to our approach to software engineering - the Virtual Engineering Organization (aka VEO).

"Is this for us?"

It's a fair question. The concept of a Virtual Engineering Organization is new enough that most people don't have a good reference point for when they make sense.

So let me be direct about it.

VEO isn't for everyone. It's not a fit if you're happy with your internal team. It's not meant to replace engineers who are already doing good work.

But there are three situations where VEO makes a ton of sense.

First: You're already outsourcing development, and it's not going well.

The problems are pretty consistent across the board. Communication is slow. Quality varies wildly. You're constantly wondering what skill level you're actually getting. When requirements evolve mid-project, it feels like starting over.

You want control over architecture and technical decisions, but you're mostly just hoping they build what you asked for.

And if you're working with offshore teams specifically? Add timezone lag to everything. Every question costs you half a day. Need someone with expertise in a specific framework? You're waiting for them to get assigned to your project, and when they do, they're starting from scratch on context.

VEO changes that equation entirely. You get spec-based development with fine-grained control. The agents have complete knowledge of every framework and language you're using. When requirements shift, there's no reassignment lag or context loss. No timezone delays. No wondering if the person working on your code actually knows what they're doing.

Second: You just raised funding and you're about to build an engineering team.

Here's the thing: the engineering team you're about to hire looks nothing like the engineering team you'll need in two years.

Right now, companies are hiring traditional engineers to write code. But in 18 months, those same companies will need operators who can manage AI agents, architects who understand how to structure work for parallel execution, and product people who can think in terms of agent orchestration rather than individual contributor capacity.

If you're building a team from scratch right now, you have a choice. Build the old model and plan to rebuild it soon. Or build an AI-first team now with the right skills and structure from day one.

The second option is harder to wrap your head around. But it's also the one that won't be obsolete before your next round of funding.

Third: You have a small internal team and you need to scale up with AI assistance.

Your engineers are good. They're using AI tools. But there's a difference between individual engineers using assistants and running an actual Virtual Engineering Organization.

One engineer with AI tools might see modest gains. Helpful, but incremental.

One operator managing a team of specialized agents? That's 8-16x throughput. And it stacks. Two agents managed by one operator is 16x. Four agents is 32x.

The constraint right now is cognitive load on the operator. Managing multiple agents doing complex technical work simultaneously is demanding. But the tools keep getting better, and operators keep getting more skilled at it.

The companies figuring this out now are the ones who'll have the advantage when the tools improve further.

So who benefits most?

Companies that want control, quality, and scale without the traditional headaches of outsourcing or massive hiring.

Companies that need engineering capacity but aren't equipped to attract and retain top-tier talent.

Companies that recognize the shift happening in software development and want to get ahead of it rather than scramble to catch up later.

Not everyone needs VEO right now. But if you're in one of those three situations, you're leaving a lot on the table by sticking with the old model.

We've been running projects this way for over a year. We've made the mistakes, learned the hard lessons, and figured out what actually works.

If you want to talk through whether VEO makes sense for your situation specifically, just reply to this email. I'm happy to be direct about whether it's a fit or not.

Here are a few other things I’ve found interesting this week, and in the case of the last one things I will be doing again and again.

1. The Adolescence of Technology by Anthropic CEO Dario Amodei. This is an incredibly insightful (and long) essay on the major risks associated with Powerful AI as we continue our march forward. The takeaways for me are that we need to be very conscious of the risks and to be thinking about ways to defend against those. I also greatly appreciate the glimpse behind the scenes at Anthropic which continues to build trust for me.

2. …But Who’s Watching the Agents? from Colin McNamara. I really like Colin’s posts - he builds AI systems for highly regulated industries so he’s clearly thought deeply about how to integrate AI and AI tools into software engineering. This post focuses on key competencies for engineering teams in the age of AI - defining clear intent, building highly observable systems, defending a very broad attack surface, and keeping a thoughtful closed feedback loop.

3. In my quest to find places to throw darts, I rediscovered the Crown & Anchor Pub just North of the UT campus. It’s literally the first bar I went to when I moved to Austin in 1995 - I lived in the apartments next door! They have a Wednesday night Cricket tournament that I’ll be frequenting from now on. See you there and save room for a burger.