Shepherds, Not Farmers

I am writing this from an orthopaedic boot and crutches after tearing my calf in half at the gym. I have not missed a day. I am limping into the office every morning because the ability to convert my thinking into multiple agents running parallel lines of work has made this the most exciting period in our company's history. The injury slowed my body. It has not slowed the business at all. A year ago, that would not have been true.

‍

In the past week alone, several CEOs of billion-dollar technology companies have published open letters to their teams announcing headcount reductions of 20% or more, framing them as the necessary response to AI reshaping how software is built. These posts are gathering millions of views. The message, repeated across companies, is consistent: AI means fewer people, flatter structures, and a new class of roles replacing the old ones.

‍

I agree with the thesis. I disagree with the method.

We did not restructure to become lean. We were built this way, and we were built this way long before AI arrived. Eventogy is a team of 28. Twelve of us build and sell the platform: eight engineers, a CTO, a product director, a commercial director, and me. We are building governance infrastructure for regulated industries. Banking. Legal. Clients who demand governance and compliance at the highest standard. We are in the middle of a complete platform rebuild, and we are not cutting. We are hiring.

Lean by choice, for fifteen years

‍

Here is the part that matters before I describe any of the tooling. We do not make economic redundancies. It is not part of how we run this company, and in fifteen years it never has been. I do not offer that as a slogan. I offer it as the way we actually operate.

‍

We do not bring people in on inflated numbers to look more serious than we are, only to let them go when the economics catch up. We do not hire for show, and we do not cut for show. We have never spent on the theatre of a modern workplace, the perks designed to look like progress, in place of the things that actually keep people. What we offer instead is longevity, real job security, and ownership of the thing you build. That is a different promise, and it attracts a different kind of person.

‍

The leanness came first. It came from being careful about who we hired and honest about why, year after year, when it would often have been easier to pad the numbers. AI did not make us lean. We were already lean. What AI did was take people who were already the right people and make them extraordinary.

‍

Engineers as shepherds

‍

Our engineers are shepherds, not farmers. They do not spend their days writing boilerplate by hand. They define the structural logic, direct AI tooling to produce the foundation, then spend their time on what actually determines quality: reviewing output, testing for regression, running code through pipelines and peer reviews. The result is three to four times faster to shippable code. Quality improves as a byproduct, because they spend more time catching problems than creating them.

‍

These are not junior developers propped up by AI. They are experienced engineers, some with us for over a decade, masters of their craft. They adopted AI tooling because that is who they are. People who stay sharp, who remain curious, who evolve with the tools available to them. There are eight of them, and they are the engine of everything that follows. The leaders I describe below set the patterns; the engineers build the product.

‍

We hire smart. We do not make redundancies. We keep people for decades, not quarters.

A CTO who designs the organisation

‍

Keiron, our CTO, designs the engineering organisation rather than just managing it. AI is embedded in every layer of the stack he oversees. He sets the patterns, the standards, and the guardrails that let eight engineers ship at the velocity of a much larger team. He runs token spend with the same discipline, so the cost of our AI leverage stays in proportion to the value it produces.

Product without a design team

‍

Graeme, our Product Director, covers product, UX, and interaction design as a single role. His workflow runs end to end in AI. Last week: concept at 5pm, working feature on dev by 10am the next morning. A two-hour working loop from idea to implementation.

‍

The prototype is the specification. Engineers read the code, run it, lift patterns directly. We shipped our Meetings Booker to sign-off without a separate design team in the loop. Not because design does not matter, but because one person with deep product instinct and the right tools covered the full cycle without compromise.

A commercial function built as a system

‍

Tom, our Commercial Director, built what he calls a commercial operating system. AI operates directly inside our CRM, email, and calendar as a connected agent. He has authored modular skill files covering every part of the commercial function: pricing, contracts, positioning, renewals, proposals, pipeline reviews.

‍

The work that a growing company would traditionally spread across sales ops, marketing, a commercial analyst, and a contracts reviewer runs through one person and the system he built. To be clear about what that means: these are roles we never needed to create, not roles we removed. The system compounds. Every interaction makes it sharper.

A CEO operating without ambiguity

‍

I record every meeting. Every call, every strategic discussion is transcribed and analysed, with everyone present notified in advance. AI surfaces signals I missed in the room. When I assess a strategic direction, the full record of how we arrived at previous decisions is documented and searchable. Patterns I did not consciously register in the moment are there to review when it counts.

‍

Ambiguity has been replaced by a trail.

Why this connects to what we build

‍

This is also why we build what we build. The way we run this company is the philosophy the product is being built to express: full oversight, nothing left to assumption, and a record that can withstand scrutiny. That principle already governs how we operate. It is what we are rebuilding Eventogy around, governance infrastructure for regulated industries, where every commercial interaction must be auditable and defensible.

What the current narrative gets wrong

‍

Here is what frustrates me about the current narrative. These open letters frame headcount reduction as vision. The language is careful, the sentiment is clear: fewer people is the point, not the side effect.

‍

And there is a pattern underneath it that nobody writing those letters wants to name. Some of the same companies now cutting in the name of AI are the ones that spent the previous years hiring for show, inflating teams and salaries to signal momentum, building the very headcount they are now congratulated for trimming. Cutting roles you should never have created is not vision. It is a correction dressed as one.

‍

We reject the whole framing. AI does not take jobs at Eventogy. It makes the people who hold those jobs more capable, more rigorous, more effective. Every person we add amplifies an already leveraged operation. We are actively recruiting even with AI embedded across every function.

‍

Hire smart people. Give them the best tools. Let them evolve. Keep them.

‍

Deliberate or forced

‍

Everything I have described is how we operate today. Not a theory. Not a roadmap. Not a manifesto. Engineers delivering at multiples of their previous velocity. A product director shipping features from concept to code in hours. A commercial director running a full department's output through a system he built himself. A CEO who has eliminated ambiguity from decision-making through disciplined recording and analysis. All while on crutches.

‍

The question for every company is not whether AI will change how they work. It will. The question is whether they design for it deliberately or find themselves forced into it.

‍

One approach builds lasting capability. The other produces headlines.

‍