I run a small DevOps and cloud-management company with six experienced engineers, and I have more than 14 years of experience across enterprise companies and startups. We've handled multi-cluster and multi-region deployments, automated and gated CI/CD pipelines, compliance preparation, monitoring, incident response, and related platform work.
Business was strong throughout the previous two years. I regularly found enough work for the team to grow, sometimes more than we could handle. Most clients were midsize companies or venture-backed startups, and we found projects through personal contacts, professional networking, freelance platforms, and our website.
That changed abruptly this year. After closing several deals per month in 2024 and 2025, I landed only two new customers in March and nothing since. Existing work gradually dried up over the summer. When I contacted former clients, many had laid off large portions of their teams, reduced infrastructure spending, or eliminated the people responsible for maintaining systems we had built.
The company account is now empty, and I may have to let the team go. I understand that AI is changing software development, but this feels like the market switched off almost overnight in early spring. Are other DevOps and cloud consultancies seeing the same thing? Is this mainly an economic slowdown, AI, companies moving past their cloud-migration phase, or something else? What would you do to adapt?
4 Answers
The market may be moving from large transformation projects to smaller, outcome-based engagements. Cloud migrations and Kubernetes rollouts have been major spending cycles for years, and many companies are now past the initial implementation phase. That does not mean infrastructure expertise is gone; it means buyers may want a narrow review, remediation project, or ongoing advisory relationship instead of a full team.
You could offer fixed-scope architecture and security reviews, incident-readiness assessments, infrastructure audits, or short consultations for startups whose products were built quickly. The revenue per client may be smaller, but the potential customer pool is much larger.
It may be less about your work suddenly becoming unnecessary and more about your customers being under pressure. When companies are cutting staff and slowing growth, consulting work such as improving pipelines or platform reliability becomes an easy expense to defer, even when it is valuable.
AI also raises the baseline for smaller teams. Developers can now generate Terraform, Helm charts, deployment scripts, and basic monitoring setups, so some companies convince themselves they can muddle through without experienced infrastructure help. The work still matters, but fewer buyers may be willing to pay for it as a general service.
A broad “DevOps services” offering may be harder to sell now. Consider packaging your expertise around a painful, measurable outcome: reducing cloud or observability bills, preparing for an audit, fixing unreliable deployments, improving security, or reviewing infrastructure created by inexperienced developers and AI tools.
One consultancy I know moved from general DevOps work into observability-cost reduction, and that created a clearer return on investment. Companies may postpone platform improvements, but they are often willing to pay when you can show that you will remove a large recurring expense or prevent a serious operational risk.
We’ve done observability work as part of larger projects, but it was rarely the main engagement. I may have underestimated how useful it could be as a focused service rather than an add-on.
There may also be an opportunity to review AI-generated infrastructure. Companies can get a prototype running quickly, but the resulting access controls, costs, reliability, and maintenance burden often need an experienced engineer.
Geography seems to matter too. Some U.S.-based startup work has slowed sharply, while European companies have remained somewhat more active in certain markets. Referrals are still powerful, but it may be worth deliberately building a referral process with former clients, engineering leaders, accountants, security firms, and software agencies in regions where spending is holding up.
I would not interpret the current pipeline as proof that your team lacks value. The combination of budget cuts, cheaper contractors, AI-assisted development, and management pressure is hitting many consultants at once. The safest response is probably to preserve cash, reduce fixed costs where possible, and reposition around problems that AI-generated or low-cost infrastructure does not solve well: accountability, security, reliability, compliance, and reducing waste.
About 80% of our previous work was with U.S. clients, 15% with European clients, and the rest elsewhere. Most of the European work came through referrals, so I need to build a more repeatable way to reach that market.
That matches what I’m seeing. Several former clients spent their investment money without reaching the growth they expected, while others were overtaken by cheaper or AI-focused competitors. Some are now trying to bootstrap with much smaller teams.