Job Placement · Technology · DevOps Engineer
A DevOps or SRE candidate evaluates your pager before your product. Google SRE publishes a ceiling of two incidents per 12-hour shift and at least 50 percent of team time on engineering rather than operations, and candidates measure your rota against it. BEG scopes that first and places permanently on Tier II milestone billing at $9,381 to $11,257.
See your placement estimate - no call required
See pricing before you talk to anyone. No demo gate, no obligation, and no co-employment.
TL;DR
Reliability candidates interview your operations practice, not the other way round. There are published numbers they compare you against: no more than two distinct incidents in a 12-hour shift, roughly six hours to handle one properly, and at least half of team time spent on engineering rather than operations. There is also a certification ladder with stated experience floors of two to three years, which tells you what a credential on a résumé is and is not evidence of. BEG scopes the rota, the platform and the credential claim before sourcing, and bills $9,381 to $11,257 against milestones rather than $19,826 to $24,783 as a percentage.
What The Candidate Is Measuring
Because the pager is the job. Published SRE practice gives candidates concrete numbers to hold your rota against, and most adverts avoid mentioning the rota at all.
The Certification Ladder, And What Each Rung Proves
Each major provider publishes an experience floor it recommends before the exam. Reading those floors turns a line on a résumé into a claim you can interrogate.
| Credential | Stated experience floor | What to ask about it |
|---|---|---|
| AWS Certified Solutions Architect, Professional | Two or more years using AWS to design and implement cloud solutions. The exam is 75 questions across 180 minutes with a 750 pass mark on a 100 to 1,000 scale, and the certification is valid for three years. | Which of those two years were production and which were a lab. Ask for one architecture decision they made that they would now reverse. |
| Google Cloud Professional Cloud Architect | Three or more years of industry experience including at least one year designing and managing solutions on Google Cloud. The exam includes two case studies. | What they migrated, and what broke. The case study format rewards people who have made trade-offs rather than people who have memorised services. |
| Microsoft Azure Solutions Architect Expert | Requires the Azure Administrator Associate credential in addition to passing AZ-305, so the certification is not awarded on the architect exam alone. | Whether they hold the certification or just the exam pass. The two are routinely conflated on résumés and it is a fair, non-hostile question. |
| No certification at all | Common among engineers who built the platform rather than studied it, and it is not a negative signal on its own. | What is running in production that they own, what the error budget is, and who gets paged when it breaks. That answer beats any credential. |
BEG treats certifications as a conversation starter rather than a filter, because filtering on them removes exactly the self-taught operators who have carried a real pager. Adjacent engineering seats: senior software engineer and IT support and help desk. Full scope on the technology placement service page.
The Fee, And Why A Percentage Punishes A Scarce Skill
Contingency pricing charges a share of first-year salary, which means the scarcer the skill the more the introduction costs you. Benchmarked to the $99,130 BLS median for network and computer systems administrators, 20 percent is $19,826 and 25 percent is $24,783. A reliability engineer will not accept that salary, though: US respondents to the 2025 Stack Overflow Developer Survey self-reported a $165,000 median total compensation for DevOps engineer and $189,000 for cloud infrastructure engineer, which puts the real contingency invoice nearer $33,000 to $47,250. BEG prices the search as a Tier II engagement at $9,381 to $11,257, agreed before sourcing and independent of what you end up paying the engineer. The fee stops being a reason to inflate the offer.
What An Understaffed Rota Costs Before Anybody Resigns
A thin on-call rotation does not fail loudly, it fails by attrition. Three people on a one-week-in-three rotation means every engineer is interrupted a third of the year, the postmortems stop being written because there is no time, and the fixes that would reduce paging never ship. Google's published six hours per incident is the reason: if the follow-up work does not fit in the shift, it does not happen, and the same incident returns. The cost shows up two quarters later as a resignation from the person who knew the most, which is a far more expensive event than the search you were avoiding.
Who The Engineer Works For
Reliability work is sold in all three shapes, and they answer different questions. The distinction that matters is who carries the pager and whose payroll they are on.
| Model | Who employs the engineer | How you pay | Right when |
|---|---|---|---|
| Managed services provider | The provider | Monthly retainer, often with an incident volume cap | You want the platform operated by somebody else entirely and accept that the operating knowledge lives outside your company. |
| Contract or staff augmentation firm | The firm | Hourly or daily rate plus margin, for the length of the engagement | A migration or a platform build has a defined end date and you do not intend to keep the headcount. |
| BEG permanent placement | You | Tier II milestone fee of $9,381 to $11,257, fixed before sourcing, against $19,826 to $24,783 on a percentage | You want the operating knowledge to stay in the building, on your payroll, with a 45-day replacement term. |
BEG is not a staffing agency and does not resell engineer time, so the retainer and the hourly rate are not products we have. The pattern worth avoiding is outsourcing the pager for two years and then discovering that nobody inside the company understands the system well enough to bring it back in. Related: director of engineering recruiting.
FAQ
A DevOps or SRE hire is a Tier II search at $9,381 to $11,257 on milestone billing. Benchmarked against the $99,130 BLS median for network and computer systems administrators, a 20 to 25 percent contingency fee would be $19,826 to $24,783, and benchmarked against what DevOps engineers actually report earning it would be considerably more.
Google SRE publishes a target of no more than two distinct incidents in a 12-hour on-call shift, on the reasoning that properly handling one incident, including root cause analysis, remediation and a postmortem, takes about six hours. If your rota routinely exceeds that, candidates will find out in the second interview and you will lose them there.
Ask for the one matching the platform you run, and treat it as evidence rather than a gate. AWS recommends two or more years of hands-on AWS design experience before the Solutions Architect Professional exam. Google recommends three or more years of industry experience including at least one on Google Cloud for its Professional Cloud Architect.
Yes, and it catches people out. Microsoft requires the Azure Administrator Associate credential in addition to passing AZ-305 before the Azure Solutions Architect Expert certification is awarded. A candidate can pass the architect exam and still not hold the certification, so read what the résumé actually claims rather than what you assume it means.
They overlap and they are not interchangeable. SRE as Google defines it is a discipline with a published budget: at least 50 percent of the team time goes to engineering work rather than operations, and teams consistently below that change their practices. DevOps as most companies advertise it has no such constraint, which is exactly what candidates are checking for.
No. A staffing or managed services firm employs the engineer and bills you a rate while they are assigned to you. BEG places a permanent DevOps engineer onto your payroll, employed by you, with no margin on their time. If what you want is someone to run the platform on a retainer, that is a managed services purchase and a different vendor.
Because the supply is narrower. Writing a service and keeping it up under load at three in the morning are different skills, and the second one is only learned somewhere that has been under load at three in the morning. Scarcity of that experience, not tooling knowledge, is what the market is pricing.
The federal data suggests the seat is migrating rather than disappearing. BLS projects network and computer systems administrators to decline 4 percent from 2025 to 2035, from 323,600 to 310,400, while software developer employment grows. The work is moving into engineering roles that write code, which is a different hire and a different pay band.
Hand over a real incident. Give the candidate a redacted postmortem from your own history with the conclusion removed and ask what they would have looked at first. That measures diagnostic instinct under ambiguity, which is the job. A quiz on Kubernetes objects measures whether they read the documentation recently.
23 to 35 days from discovery call to signed offer on average, based on isolved placement data, with an 86 percent fill rate on active searches. The most common delay is internal: nobody has written down what the on-call rota actually looks like, and candidates will not accept without it.
US respondents to the 2025 Stack Overflow Developer Survey reported a median total compensation of $165,000 for DevOps engineer or professional, with cloud infrastructure engineer at $189,000. That is self-reported survey data rather than BLS, so treat it as a market signal to sanity-check your band against rather than as an official wage estimate.
A 45-day replacement guarantee and 50 percent off a repeat search for the same seat. The honest caveat is that reliability hires who leave early usually leave because the on-call load turned out heavier than described. Fixing the rota matters more than replacing the person, and we will say that before we open a second search.
Building out the platform team? See software engineer, senior engineer and tech lead, IT support and help desk and CTO and CIO, or all technology placement.
More technology placement
Ready?
Answer a few questions, get your placement estimate in about 90 seconds. No call required, no commitment.
See your placement estimate - no call required
