Dmitry Saksonov on the Cost of Growing Too Fast
October 1, 2026 · #atom group #leadership #crisis
Contents
- Why I’m writing this
- The 2014 decision
- Rebuilding the team
- What happened in 2024 and 2025
- How the company operates in 2026
Why I’m writing this
I’m building ATOM, a technology group developing applied AI products across several fields. Some of those products have already launched in the U.S. market. Anyone considering working with me can review the facts and timeline or read the questions and answers.
At fourteen, I started working on construction sites near Kolodishchi to earn money for a motorcycle. I organized my first crew there.
The 2014 decision
I had already spent several years in the automotive industry, where I started out in auto repair.
In 2014, I used borrowed money to buy cars for resale. While the cars were in transit, the import and sales rules changed. We had to suspend sales for several months, during which the exchange rate collapsed and the market weakened. That was my mistake: I planned as if conditions would not change and left myself no cushion. I spent the next two years dealing with the consequences.
Four years later, this became the basis for a criminal case against me. I was sentenced in 2018. While the case was being reviewed, I remained in a pretrial detention center and was never transferred to a prison or penal colony. The sentence was later revised; I was released early, and my criminal record was subsequently expunged. The court order confirming that is published below.
Rebuilding the team
When I resumed work, I brought together a group that stayed with me for years and formed the backbone of the management team. They gave us continuity as the company changed direction.
We first returned to what we already knew: cryptocurrency mining. The model depended on hardware assembly, hosting, and physically deployed infrastructure, so an external disruption could bring the entire operation to a halt. Once its vulnerability became clear, we began shifting from physical infrastructure to digital products and eventually moved into software and applied AI.
What happened in 2024 and 2025
In 2024, the company hired rapidly to fulfill a large contract for a single client. We paid above-market salaries to deliver faster, including on weekends. At its peak, the team reached 1,500 people.
As the contract required more capacity, we continued adding people. Each round addressed the immediate workload but drove monthly costs higher. We built payroll around the client’s payments without leaving an adequate buffer.
I approved that pace. The business grew faster than we could manage or afford, and too much of our revenue depended on one client. Those decisions were mine. When those payments stopped in late 2024, I used my own money and took on additional debt to keep the company operating.
For a while, those measures covered payroll and bought us time, though revenue still fell short of monthly costs. At that scale, even a brief delay created a gap that widened with every payroll cycle. Waiting for one payment would not solve it.
Debt grew, salaries were delayed, and people left. By 2025, cuts were unavoidable, even though they would not settle the existing obligations.
At the time, I described the company as a plane that had run out of fuel with people still on board. We stayed aboard and landed it.
We reduced the team to a size the business could support. We stopped work that current revenue could not fund and rebuilt the plan around actual cash flow. The goal was to continue delivering while bringing costs to a sustainable level.
I had said publicly that I would not walk away from the problem. We had to pay what the company owed to people without shutting it down.
I went abroad to seek financing. Once it was secured, the company continued operating and completed the remaining payments.
By 2026, we had settled all of those obligations. I returned, and the company resumed growth with its core team in place. Before adding people, we changed how we made hiring decisions. Those rules now apply throughout the business.
How the company operates in 2026
Cash comes before headcount. Every hiring plan starts with funds already on hand; projected revenue does not count. Before approving a role, we test whether we could keep paying the salary if revenue fell or a major payment was delayed. A plan that requires new borrowing does not go forward.
The company cannot depend on a single client or partner. We track both the source and the volume of revenue. Before we let a large contract change our cost base, we ask whether the business could keep operating if payments stopped. We do not build a permanent team on the strength of one contract.
Leadership capacity has to be in place before the team expands. We assess whether current managers can lead a larger group and whether each decision has a clear owner. Day-to-day work must move forward without constant escalation to the founder. We do not hire until the structure is ready.