Hire Cloud, DevOps & SRE Engineers
The roles that decide whether everything else ships. We screen for production ownership rather than for certifications, because in this family the CV and the capability diverge more than anywhere else in engineering.
Which of these do you actually need?
More briefs go wrong here than at any other point. These four titles appear interchangeable in job adverts and are not interchangeable in practice.
DevOps engineer
Hire when: Releases are slow, environments are inconsistent, deployment is manual.
Owns pipelines, build and release tooling, environment provisioning. Measured on delivery velocity and repeatability.
Site reliability engineer
Hire when: Things break, recovery is slow, nobody owns uptime.
Owns the running system: monitoring, error budgets, on-call rotation, incident response and post-incident work. Measured on availability, not throughput.
Platform engineer
Hire when: Your DevOps team has become the bottleneck for every other team.
Builds internal tooling and self-service infrastructure so product teams ship without filing tickets. A product role whose customers are your own engineers.
Cloud architect
Hire when: A migration, a multi-cloud decision, or a cost problem nobody can explain.
Designs the target state — topology, networking, security boundaries, cost model. Usually a smaller, more senior engagement than the three above.
How we screen infrastructure people
Every CV in this family lists the same eight tools. Separating the people who built and owned from the people who used what others built takes specific questions, not a longer keyword list.
What we need in the brief
A precise brief on these roles removes about a fortnight from the search.
- Which cloud, and whether multi-cloud is real or aspirational.
- Scale in terms an engineer recognises: services, environments, deploys per week.
- Whether the role carries on-call, and what the rotation looks like.
- Whether this is building something new or running something existing.
- Which tools are genuinely required and which are habit from the last hire.
Questions employers ask
What is the difference between a DevOps engineer and an SRE?
In Indian job adverts the titles are used almost interchangeably, which is why briefs so often produce the wrong shortlist. In practice a DevOps engineer is usually measured on delivery — pipelines, environments, release velocity — while a site reliability engineer is measured on the running system, with error budgets, on-call and incident response as the core of the job. If your pain is that releases are slow, you want DevOps. If your pain is that things break and nobody owns it, you want SRE, and hiring a DevOps engineer will not fix it.
Is a platform engineer the same thing?
No, and the distinction matters at scale. Platform engineers build the internal tooling that other engineers use to ship — the paved road, the self-service environments, the abstractions over Kubernetes. It is a product role whose users are your own developers. Organisations usually need it once they have enough engineering teams that a central DevOps function becomes a bottleneck, generally somewhere past forty or fifty engineers.
Why are Kubernetes and Terraform skills so hard to verify?
Because both are claimed far more often than they are held at production depth. Large numbers of candidates have used a cluster someone else built and run modules someone else wrote. The people who can design cluster topology, reason about networking and failure modes, or own infrastructure-as-code across environments are a much smaller pool. Screening has to reach past tool names into decisions the candidate personally made and incidents they personally handled — CV keywords will not separate them.
What does cloud and DevOps talent cost in India?
The live band across our open mandates is shown in the panel on this page and updates as roles open and close, so it reflects what clients are paying now rather than what a salary report said last year. As a general shape: the premium sits with production ownership and multi-cloud depth rather than with certifications, and a candidate who has genuinely run a platform at scale commands materially more than one who has operated within someone else’s.
Should we hire cloud engineers permanently or on contract?
Migrations argue for contract, steady-state argues for permanent. A cloud migration or a Kubernetes adoption is a defined project with an end, and contract specialists who have done it several times will finish faster than a permanent hire learning on your infrastructure. Running the platform afterwards is continuous work with accumulated context, which is a permanent role. A common and effective pattern is contract specialists for the migration alongside a permanent engineer who absorbs the knowledge.
Which Indian cities have the deepest cloud and DevOps pools?
Bangalore, Hyderabad and Pune, with Delhi NCR strong at the enterprise and managed-services end. The live panel on this page shows where our own open mandates sit today. Remote hiring genuinely widens the pool for these roles more than for most — infrastructure work is less dependent on physical presence than almost anything else in engineering.
Hiring for infrastructure?
Tell us what is breaking or what is slow. We will tell you which of the four roles you actually need.