2026-07-01
DevOps / SRE: 2,498 open roles, $187.5k median — AI is making infrastructure more complex, not simpler
2,498 open DevOps and SRE roles. 1,088 companies. Median $187.5k USD. The DevOps/SRE market in July 2026 — from live ATS data.
Every new LLM deployment is a new reliability problem. GPU orchestration, inference latency SLOs, vector database maintenance, model serving autoscaling — the AI era has extended the scope of DevOps/SRE rather than automated it away. The people who own infrastructure reliability are in higher demand, not lower.
I track DevOps Engineer and SRE roles daily from public ATS feeds. Here's July 2026.
What DevOps and SRE engineers do
DevOps Engineers build and maintain the infrastructure, CI/CD pipelines, and tooling that let software teams ship reliably and frequently. Kubernetes, Terraform, GitHub Actions, Helm, cloud provider APIs — the DevOps stack is the invisible layer that makes everything else work.
Site Reliability Engineers (SREs) own the reliability contract: uptime, latency, error budgets, incident response. Where DevOps is about shipping speed, SRE is about production stability. The two roles overlap heavily in practice — most modern organisations use the titles interchangeably or employ generalists who cover both.
The AI infrastructure dimension: ML model serving has unique reliability challenges — GPU resource contention, model cold starts, inference timeouts at scale, embedding pipeline latency. Senior DevOps/SRE engineers with GPU infrastructure experience are a genuinely scarce skill set in 2026.
The data: 2,498 open roles across 1,088 companies
2,498 active roles. 1,088 companies. 330 new roles in the last 7 days.
1,088 companies is the widest company base of any niche we track — narrowly edging out data engineering. DevOps/SRE is the most universally needed technical discipline: every company with software in production needs someone ensuring it stays up.
Who's hiring most aggressively
| Company | Open DevOps/SRE roles |
|---|---|
| Accenture Federal Services | 53 |
| NVIDIA | 53 |
| Cisco | 36 |
| Air Apps | 33 |
| Okta | 27 |
| Qube Research & Technologies | 22 |
| ClickHouse | 20 |
Accenture Federal Services and NVIDIA tied at 53 — both are in aggressive scale-up mode. Accenture Federal is running large government cloud migration programs that require extensive DevOps engineering. NVIDIA's infrastructure team is scaling to support both their internal AI compute needs and the DGX Cloud platform customers.
ClickHouse at 20 is notable — a real-time analytics database company hiring at this density signals rapid product-led growth and significant infrastructure demands from enterprise customers.
Qube Research & Technologies (quantitative hedge fund) hiring 22 DevOps/SRE roles signals that financial firms are now competing directly with tech companies for infrastructure engineering talent, at tech-company compensation levels.
What they pay
Of 2,498 roles, 763 (30.5%) have a published salary range. Among USD roles:
- Median: $187,500/year
- 25th percentile: $156,600
- 75th percentile: $225,000
The $68k P25-P75 spread reflects the wide range of DevOps work — from CI/CD automation at mid-market SaaS to GPU cluster management at an AI lab. The $187.5k median is strong and has been appreciating steadily, driven by cloud complexity growth and AI infrastructure demands.
Freshness: 330 new roles in 7 days
330 per week is consistent with a large, stable market. Not the explosive growth of ML engineering or agentic engineering — but DevOps/SRE is the most durable of the technical disciplines. It doesn't go through boom/bust hiring cycles because production systems never stop needing reliability engineers.
What this tells us about the DevOps/SRE market
1. 1,088 companies makes this the most broadly distributed technical hiring market. You can find DevOps/SRE roles at fintech, pharma, defence, AI startups, cloud platforms, gaming companies. The breadth means you can optimise for sector, company size, or domain interest without limiting your options.
2. Federal infrastructure is an underrated hiring magnet. Accenture Federal Services at 53 roles is the top hirer in this market. Federal DevOps often involves security-hardened infrastructure (FedRAMP, DoD environments), interesting scale challenges, and comp bands that have been catching up to commercial tech. Clearance-eligible candidates have a material advantage.
3. NVIDIA's infrastructure buildout is both opportunity and signal. NVIDIA building a 53-person DevOps/SRE presence isn't just an employment opportunity — it's a signal that AI compute infrastructure has become complex enough to require a dedicated platform engineering organisation at the company that makes the chips. The expertise being built there will be highly portable.
4. The AI premium for GPU infrastructure experience is real. The median is $187.5k for all DevOps/SRE. Engineers with Kubernetes GPU scheduling, CUDA-aware infrastructure, or ML serving experience (Triton, vLLM, Ray Serve) are quoting $220k–$280k+ at frontier AI labs. The AI infrastructure sub-specialty is currently undersupplied.
The board
latchhire.com/board/devops_sre — updated daily. Every role links to the original posting.
Want new DevOps / SRE roles weekly? Subscribe to job alerts →
Data pulled 2026-07-01. Active roles only. Salary USD only. "New in 7d" = first seen in the past 7 days.