A Staff Site Reliability Engineer at Google, Meta, or Stripe earned a median $560,000 in total compensation in H1 2026 -- base, annualised equity, and target bonus combined. That figure, drawn from Levels.fyi public submissions, the Radford Technology Compensation Survey, and the ENTRA Salary Survey IT infrastructure panel, represents an 18% increase over H2 2025 and places the Staff SRE in the same compensation tier as VP-level engineering leadership at mid-market technology companies. The driver is structural and AI-specific: every hyperscaler and large-scale platform is now operating AI inference workloads alongside traditional serving infrastructure, and the reliability engineering demands of LLM serving at scale have made the Staff SRE one of the most structurally critical and correspondingly expensive engineering roles in the market. This index maps the full IT infrastructure pay stack -- from the $560K Staff SRE floor at Big Tech to the $175K Senior DevOps Engineer at enterprise employers -- and captures the AI-adjacency premium that is now the dominant comp variable across the cloud, platform, MLOps, and SRE role families.
The IT infrastructure pay stack
All figures represent median total compensation: base salary + annualised equity (4-year vest) + target bonus. P75 is the 75th-percentile total comp, capturing above-median earning potential. Demand heat is scored against the 30-day rolling average of open postings as of June 2026. YoY delta is computed against ENTRA H2 2025 IT Infrastructure Survey and Radford H2 2025 Technology benchmark. All figures USD.
| Rank | Role | Score | Rating | Median Total Comp | P75 | YoY | |---|---|---|---|---|---|---| | 1 | Staff Site Reliability Engineer (Google / Meta / Stripe) | 97 | AAA | $560,000 | $720,000 | +18% | | 2 | Staff ML Infrastructure Engineer (Frontier Lab) | 95 | AAA | $520,000 | $680,000 | +26% | | 3 | Staff MLOps Engineer (AI Lab) | 93 | AA+ | $490,000 | $630,000 | +24% | | 4 | Principal Cloud Architect (Big Tech / Hyperscaler) | 91 | AA+ | $460,000 | $590,000 | +15% | | 5 | Principal Platform Engineer (AI-Native Company) | 89 | AA | $430,000 | $560,000 | +22% | | 6 | Staff Security Engineer - AI Systems (Frontier Lab) | 87 | AA | $410,000 | $540,000 | +20% | | 7 | VP of Engineering - Infrastructure (Tech Scale-up) | 85 | AA | $390,000 | $510,000 | +14% | | 8 | Senior Staff Data Engineer (Big Tech) | 83 | A+ | $380,000 | $490,000 | +16% | | 9 | Principal DevOps Engineer (Cloud Provider) | 81 | A+ | $360,000 | $470,000 | +13% | | 10 | Senior Cloud Security Architect (Regulated Enterprise) | 79 | A | $335,000 | $440,000 | +17% | | 11 | Platform Engineering Manager (Growth-Stage Startup) | 77 | A | $310,000 | $410,000 | +15% | | 12 | Senior MLOps Engineer (Enterprise AI) | 75 | A | $285,000 | $370,000 | +21% | | 13 | Senior SRE (Fintech / Neobank) | 73 | BBB | $265,000 | $345,000 | +12% | | 14 | Principal Database Reliability Engineer | 71 | BBB | $255,000 | $335,000 | +10% | | 15 | Senior Data Reliability Engineer (AI Platform) | 69 | BB | $245,000 | $320,000 | +18% | | 16 | DevOps Architect (Healthcare Tech) | 67 | BB | $230,000 | $300,000 | +13% | | 17 | Senior Kubernetes / Cloud-Native Engineer | 65 | BB | $215,000 | $280,000 | +9% | | 18 | Senior Platform Engineer (AI Startup) | 63 | BB | $200,000 | $270,000 | +14% | | 19 | Site Reliability Engineer L5 (Mid-size Tech) | 61 | BB | $195,000 | $255,000 | +8% | | 20 | Senior DevOps Engineer (Enterprise) | 59 | B | $175,000 | $225,000 | +7% |
Top 10 roles in detail
#1 -- Staff Site Reliability Engineer (Google / Meta / Stripe)
The Staff SRE is the highest-compensated IT infrastructure individual-contributor role in this index, and H1 2026 widened the gap to second place. Median total comp of $560K is composed of a base of $220K-$280K, annualised RSU grants of $260K-$380K, and a cash bonus of $40K-$80K. The p75 of $720K reflects equity refreshes granted to engineers on-call for revenue-critical AI inference products. The structural driver is the expansion of SRE scope into AI reliability: Staff SREs at Google, Meta, and Stripe are now responsible for reliability engineering on LLM inference pipelines in addition to traditional serving infrastructure. That expansion -- latency spikes from variable prompt lengths, GPU memory OOM cascades, autoscaling under unpredictable load distributions -- requires a new class of SRE expertise that the existing candidate pool has not yet fully developed, keeping supply at extreme scarcity. LinkedIn Talent Insights data, per ENTRA analysis, shows fewer than 900 active candidates at this credential level across all three employers combined, against over 300 open Staff SRE requisitions (ENTRA estimate, LinkedIn member + posting data, June 2026). YoY movement of +18% is sustained by this supply-demand imbalance with no near-term correction visible.
#2 -- Staff ML Infrastructure Engineer (Frontier Lab)
Staff ML Infrastructure Engineer at a frontier lab recorded the largest YoY comp movement in this index at +26%, reaching $520K median. The role owns the distributed training infrastructure that frontier-lab research teams run on: job scheduling across GPU clusters of 4,000-64,000 accelerators, fault-tolerant training resumption, custom CUDA kernel integration, and the tooling that allows researchers to iterate on architecture experiments without waiting for infrastructure teams. Anthropic, OpenAI, Google DeepMind, Meta AI, and xAI all announced major compute capacity expansions in H1 2026, and each new cluster requires Staff ML Infra Engineers to commission, validate, and operate it. The supply pool is the thinnest in this index in its credential specificity: the intersection of GPU cluster management, custom interconnect knowledge (InfiniBand, NVLink topology design), and ML framework depth (JAX, PyTorch distributed, Megatron-LM) is a population of under 500 engineers globally who meet the Staff bar, per ENTRA estimate based on LinkedIn credentialed-member filter and ENTRA recruiter survey data (n=52, H1 2026). The comp includes a base of $210K-$260K, large equity refreshes on 12-month cycles at several frontier labs, and retention bonuses that have become standard for incumbents with active competing offers.
#3 -- Staff MLOps Engineer (AI Lab)
Staff MLOps Engineer at an AI lab or AI-native company reached $490K median in H1 2026, recording a +24% YoY movement driven by AI-adjacency demand rather than traditional IT infrastructure supply dynamics. The role owns the production ML lifecycle end-to-end: model versioning and registry, training pipeline orchestration (Kubeflow, Metaflow, Airflow variants), feature-store management, model serving infrastructure (Triton, vLLM, TensorRT-LLM), and the observability stack that monitors model performance in production. The AI-adjacency premium is the dominant comp variable at this level: Levels.fyi data shows Staff MLOps Engineers at AI labs earning a 28% premium over equivalent Staff Platform Engineers at companies without ML production workloads. Demand heat is extreme -- LinkedIn posting velocity for Staff-equivalent MLOps roles grew 42% in H1 2026 (per LinkedIn Talent Insights, ENTRA analysis, H1 2026 vs H1 2025), the second-largest posting growth of any role family in this index. The comp structure at frontier labs for this level includes a base of $190K-$240K, annualised equity of $220K-$360K, and a cash bonus of $40K-$60K.
#4 -- Principal Cloud Architect (Big Tech / Hyperscaler)
Principal Cloud Architects at AWS, Google Cloud, Microsoft Azure, and Oracle Cloud cleared $460K median, driven by the $500B+ collective hyperscaler capex commitment to AI data centre buildout in H1 2026. The role is the technical design authority for multi-region, multi-AZ architectures that govern how AI training and inference workloads run on cloud infrastructure -- GPU cluster networking, high-bandwidth storage tiering for model checkpoints, and inference autoscaling at sovereign scale. Comp at this level reflects the combination of deep technical authority and tenure-inflated equity: base $200K-$250K, annualised RSUs of $200K-$300K (reflecting equity grants accumulated over 6-10 years at hyperscalers), and a bonus of $40K-$80K. New hires at Principal Cloud Architect level command sign-on packages of $200K-$400K to offset unvested equity at prior employers. The supply side is constrained not by raw scarcity but by experience depth: Principal Architects require demonstrable experience designing for petabyte-scale AI workloads, a credential that only became achievable in the last three to four years.
#5 -- Principal Platform Engineer (AI-Native Company)
Principal Platform Engineers at AI-native companies -- Anthropic, OpenAI, Cohere, Scale AI, Anyscale, Mistral, and comparable organisations -- reached $430K median with a +22% YoY movement. At these organisations, the internal developer platform is the substrate on which all model development, evaluation, and deployment runs. A Principal Platform Engineer's decision about orchestration framework standardisation or internal compute-scheduling API design has direct bearing on research team velocity and therefore on competitive trajectory. The comp structure is equity-dominant in a way that is distinct from hyperscaler RSUs: AI-native companies offer a mix of RSUs (at organisations with liquid equity) and pre-IPO options (at private companies), where the options at current funding-round strike prices represent 5-10x implied upside on a successful outcome. The $430K median valued at current round prices may significantly understate or overstate the actual realised value depending on exit trajectory. Demand heat is extreme: approximately 14 new companies in this employer cohort raised Series B or C rounds in H1 2026 (ENTRA estimate based on Crunchbase funding round tracking, H1 2026), all competing for the same thin supply of Principal-level platform talent.
#6 -- Staff Security Engineer - AI Systems (Frontier Lab)
Staff Security Engineer specialising in AI systems at a frontier lab reached $410K median, driven by the rapid formalisation of AI-specific security as a dedicated practice at Anthropic, OpenAI, Google DeepMind, and Meta AI. The credential intersection that this role requires -- offensive security depth plus AI-systems knowledge sufficient to model attack surfaces unique to neural network architectures -- is the rarest combination in the IT security market. The specific threat vectors that this role addresses are not covered by traditional application or network security frameworks: prompt injection at the API layer, training-data poisoning at the data-collection stage, model-weight exfiltration, and adversarial inputs designed to bypass safety classifiers. All four major frontier labs have active hiring at the Staff level for this specialisation, with requisitions often sitting open for six months or more. The comp of $410K median -- base $200K-$260K, annualised equity $170K-$250K, bonus $30K-$60K -- reflects the scarcity premium on a credential combination that the security market has had fewer than three years to develop.
#7 -- VP of Engineering - Infrastructure (Tech Scale-up)
VP of Engineering - Infrastructure at a tech scale-up cleared $390K median in H1 2026, representing the first management-layer entry in this index. The role has become the first VP-level engineering hire at many AI-native companies scaling from prototype to production at 10-100x traffic growth: the VP of Infrastructure is the person who decides whether that scaling happens cleanly or catastrophically. Comp of $390K median includes base $250K-$310K, options annualised at $100K-$180K at current round valuations, and a cash bonus of $40K-$70K. The management premium over the Senior Staff IC roles below it in this index is moderate -- 3% above Senior Staff Data Engineer -- reflecting the market's current valuation of IC technical depth versus management span at the infrastructure layer. At growth-stage companies specifically, the VP of Infrastructure role is often the most difficult senior engineering hire: it requires a track record of scaling distributed systems from 10M to 100M+ daily requests AND people leadership experience AND the operational judgement to make infrastructure investment decisions with direct cost implications on a lean pre-revenue budget.
#8 -- Senior Staff Data Engineer (Big Tech)
Senior Staff Data Engineers at Big Tech reached $380K median, with YoY growth of +16% driven by the AI training-data imperative. At Google, Meta, Microsoft, Amazon, and Apple, training data quality is now a direct competitive variable in the AI model quality race, and Senior Staff Data Engineers who design petabyte-scale pipelines that clean, deduplicate, curate, and route training datasets are doing model-quality-determining engineering rather than analytics or reporting work. The structural repricing of this role versus its historical compensation band reflects this shift: Senior Staff Data Engineers at Big Tech now earn within 5% of Senior Staff ML Engineers at the same companies, a gap that was 20-25% as recently as 2023, per ENTRA estimate based on Levels.fyi historical band comparison and Radford H2 2023 Technology Compensation Survey. The comp of $380K includes base $190K-$240K, annualised RSUs of $130K-$220K, and a bonus of $30K-$60K. All five Big Tech employers increased data engineering headcount in the AI training function by double-digit percentages in H1 2026 relative to H1 2025.
#9 -- Principal DevOps Engineer (Cloud Provider)
Principal DevOps Engineers at cloud providers reached $360K median with a +13% YoY movement, as the baseline requirement for the role expanded to include AI workload CI/CD pipelines: ML model artifact management, Kubernetes cluster provisioning for heterogeneous GPU and CPU node pools, and multi-region deployment orchestration for AI inference services with strict latency SLAs. The comp of $360K -- base $170K-$220K, annualised RSUs of $130K-$200K, bonus $30K-$50K -- reflects the combination of cloud-provider tenure-inflated equity and the expanded technical scope. LinkedIn posting velocity for Principal DevOps roles at cloud providers grew 31% in H1 2026 as every major cloud provider scaled its AI service engineering teams. The supply side is less constrained than the Staff-level roles at the top of this index, but the AI-era specificity of the requirements -- GPU node management, Triton server deployment pipelines, cost-allocation tagging for heterogeneous compute -- has narrowed the qualified candidate pool relative to the historical Principal DevOps market.
#10 -- Senior Cloud Security Architect (Regulated Enterprise)
Senior Cloud Security Architects at regulated enterprises reached $335K median in H1 2026, recording a +17% YoY movement that is the largest comp increase for any role below the Staff IC tier in this index. The structural driver is external and legislative: the EU AI Act's requirements for high-risk AI systems -- including data-residency controls, access-logging obligations, and security testing mandates -- created a demand wave for architects who understand both cloud security frameworks (SOC 2, ISO 27001, FedRAMP) and the AI-specific obligations in the EU AI Act's Annex III categories. At regulated enterprises in financial services, healthcare, insurance, and critical infrastructure, the Senior Cloud Security Architect is now accountable for validating that AI system deployments comply with a security architecture standard that combines existing cloud-security requirements with new AI-regulatory obligations. The comp of $335K reflects base $170K-$210K, bonus $30K-$60K, and regulated-enterprise benefits packages that add $20K-$40K in effective value. The credential requirement is the primary scarcity driver: cloud security architects with AI Act compliance experience are a population measured in hundreds, not thousands, globally.
What the data is telling us
The defining structural finding of the IT Cloud and DevOps Salary Index H1 2026 is the bifurcation of the IT infrastructure market along a single axis: AI adjacency. The same job title at an AI-adjacent employer commands a materially different compensation band from the equivalent title at a non-AI employer -- and the premium is growing rather than converging. Staff MLOps Engineers at AI labs earn 28% more than equivalent Staff Platform Engineers outside the ML stack. Staff ML Infrastructure Engineers at frontier labs earn 26% more YoY, while Senior Kubernetes generalists grew 9%. Principal Platform Engineers at AI-native companies earn $430K; equivalent Principal Platform Engineers at legacy enterprise employers not in this index earn $220K-$260K. The premium is not a rounding error -- it is a structural market signal that the AI infrastructure buildout has created a two-tier IT labour market, and the tier differential is widening.
The second major finding is the formalisation of new engineering disciplines at above-market rates. Staff ML Infrastructure Engineer, Senior MLOps Engineer, and Senior Data Reliability Engineer are all roles that did not have standard levelling frameworks or robust comp data as recently as 2023. In H1 2026 all three are benchmarkable, heavily posted, and growing faster than established IT infrastructure disciplines. When a discipline formalises faster than the labour supply can respond -- when there are more open roles than credentialed candidates who have had the two to three years of production experience required to be competitive at Senior or Staff level -- the comp accelerates. This is the mechanism behind the +24% for Staff MLOps, the +21% for Senior MLOps, and the +18% for Senior Data Reliability. It will moderate once the supply pipeline catches up; the H2 2026 index will be the first data point on whether that moderation has begun.
The most underpriced role in this index is the last one. A Senior DevOps Engineer at an enterprise organisation earns $175K median -- a competitive, durable wage for a role that is genuinely valuable -- but the supply pool is not scarce, the role definition is heterogeneous, and the AI-adjacency premium does not reach this title at the generalist level. The engineers who want to accelerate their comp trajectory out of this band have a clear path in H1 2026: develop credible production MLOps or ML infrastructure experience. The distance between a Senior DevOps Engineer at $175K and a Staff MLOps Engineer at $490K is primarily a credential gap, not a tenure gap. The market is communicating this clearly in H1 2026. The engineers who hear it will drive the supply-side response that moderates the upper band of this index in 2027 and beyond.
How we ranked
The Top 20 IT Cloud and DevOps Engineering Salary Index H1 2026 is scored across 4 dimensions:
- Median Total Compensation (40%) -- median base + annualised equity (4-year vest) + target bonus (Source: Levels.fyi public submissions for SRE, Platform, Cloud, DevOps, and MLOps titles Jan-Jun 2026, n=2,400+; Radford Technology Compensation Survey H1 2026 infrastructure engineering cut; Glassdoor verified salary reports for DevOps and SRE; BuiltIn.com cloud engineering salary surveys; ENTRA Salary Survey H1 2026, IT infrastructure respondents, n=184)
- P75 Total Compensation (20%) -- 75th-percentile total comp per role, capturing realistic above-median earning potential (Source: Levels.fyi p75 band; Radford infrastructure engineering cut; ENTRA Salary Survey H1 2026, IT infrastructure respondents, n=184)
- Demand Heat (20%) -- 30-day rolling average of open postings across LinkedIn, Indeed, and Greenhouse as of June 2026, normalised for title-level scarcity and time-to-fill (Source: LinkedIn Talent Insights June 2026; ENTRA recruiter survey n=52 technical recruiting professionals H1 2026)
- Supply Scarcity (20%) -- estimated credentialed talent pool depth: LinkedIn member count at qualifying title + seniority + cloud-domain filter, cross-referenced against Radford supply-side data. Scored 1-10 inverted -- scarcer supply = higher score (Source: LinkedIn member data; Radford Workforce Analytics; ENTRA recruiter survey H1 2026)
Data window: January 1 - June 20, 2026 Sample size: 2,400+ Levels.fyi SRE/Platform/Cloud/DevOps/MLOps submissions; Radford Technology Survey H1 2026 infrastructure cut; 184 verified ENTRA Salary Survey IT infrastructure respondents; LinkedIn Talent Insights posting-velocity data for 20 role families Year-over-year delta: computed against ENTRA Salary Survey H2 2025 IT infrastructure cut and Radford H2 2025 Technology Compensation benchmark.
Limitations:
- Equity figures for roles at private AI labs and growth-stage startups are estimated from disclosed funding-round valuations; actual realised value at liquidity may differ materially from the reported medians.
- Staff and Principal-level roles at hyperscalers (Google, Meta, Amazon) dominate the Levels.fyi sample for the top tier of this index; mid-market and enterprise figures rely more heavily on Radford and ENTRA survey data, which may under-represent outlier compensation at smaller high-growth employers.
Inquiries about methodology: methodology@entracareers.com
Year-over-year movement
The most significant year-over-year shifts in H1 2026 IT infrastructure compensation are concentrated in three role families.
Staff ML Infrastructure Engineer (+26%) is the largest single-role YoY movement in this index and one of the largest in the full ENTRA H1 2026 salary dataset. The driver is the pre-training compute race: every major frontier lab expanded its GPU cluster capacity in H1 2026, and every new cluster requires Staff-level ML infra engineers to commission and operate. The supply response to this demand signal is slow -- the credential combination required at the Staff level in this role takes four to six years of specialised experience to develop -- and no significant supply-side correction is visible in the H1 2026 data.
Staff MLOps Engineer (+24%) and Senior MLOps Engineer (+21%) reflect the AI production deployment wave hitting both the frontier-lab and enterprise markets simultaneously. LinkedIn posting velocity for MLOps roles grew 42% (Staff) and 68% (Senior) in H1 2026 relative to H1 2025, per LinkedIn Talent Insights data reviewed by ENTRA. The Senior MLOps number is the most surprising in this index: a 68% posting-velocity increase is a demand signal that typically precedes a comp correction larger than the +21% already recorded. The H2 2026 index is likely to show further Senior MLOps acceleration as the Fortune 500 AI production cohort expands.
Senior Cloud Security Architect (+17%) is the standout movement outside the direct AI-infrastructure stack, driven entirely by the EU AI Act's enforcement cycle. This role will remain an accelerating comp story through H2 2026 as the August 2026 high-risk AI system obligations come into force for companies with EU operations.
Roles that are decelerating relative to prior periods: Senior DevOps Engineer (+7%) and SRE L5 at mid-size tech (+8%) reflect a normalising mid-market band after the broad infrastructure compensation reset of 2022-23. The AI-adjacency premium does not reach these generalist titles, and the supply pool at Senior level has caught up with demand.
What's next for IT infrastructure pay
The H2 2026 outlook for IT Cloud and DevOps compensation is defined by two forces pulling in opposite directions.
On the demand side, the AI infrastructure buildout is accelerating through H2 2026. All four major hyperscalers have confirmed Q3 and Q4 data centre commissioning milestones that will require additional Staff-level ML Infrastructure, SRE, and Platform Engineering headcount. The frontier-lab tier has expanded: three additional well-funded labs (two US-based, one European) are expected to reach training-cluster scale in H2 2026, adding to the demand pool for Staff ML Infrastructure Engineers and Staff MLOps Engineers. The enterprise AI production wave is in its early innings: fewer than 15% of Fortune 500 AI initiatives that were in pilot in H1 2026 have reached full production deployment. As that percentage grows through H2, the demand for Senior MLOps Engineers and Senior Data Reliability Engineers at enterprise employers will continue to expand.
On the supply side, the first cohort of engineers who upskilled specifically into MLOps and ML infrastructure in 2023-24 is now reaching Senior-level experience thresholds. The Radford workforce analytics data shows a 34% increase in LinkedIn profiles updating to MLOps or ML infrastructure titles in H1 2026, suggesting the supply response to the demand signal is beginning. This will not close the Staff-level gap in H2 2026 -- you cannot accelerate four to six years of experience -- but it will begin to moderate the Senior MLOps and Senior Platform Engineering comp trajectory from the +21-24% range into the +12-16% range by H2 2026.
The roles to watch for the next index refresh: Staff Security Engineer - AI Systems, where the EU AI Act's conformity assessment obligations and the US Administration's AI security framework are likely to create a second demand wave in H2 2026; and Senior Data Reliability Engineer, where the enterprise AI data-quality incident rate is accelerating and the supply-side response has not yet begun. Both are candidates for the largest YoY movements in the H2 2026 IT infrastructure index.
Global Career PlatformFind AI talent. Find your next role.
Booking is hotels. · Airbnb is apartments. · ENTRA is global careers.