Who & why
Keeps the product running in prod and shipping fast — replacing manual toil with automation and making reliability a measurable engineering task. DevOps = delivery culture & automation (CI/CD, IaC); SRE = reliability via metrics (SLO/SLI, error budget). Without one: fragile infra, manual erroneous deploys, no monitoring (users tell you it's down), no backups/recovery, chaotic repeating incidents.
A day in the life
Morning: check dashboards/alerts, SLOs, error budget; triage night alerts. Day: improve CI/CD, write IaC, set up monitoring for a new feature, coordinate incident response. Evening: write a runbook, a blameless postmortem, verify backups & recovery, automate what was manual.
Key skills
Hard: Linux, networking, containers (Docker), orchestration (K8s), CI/CD, IaC (Terraform/Ansible), cloud, monitoring/logs (Prometheus/Grafana/Sentry), SLO/SLI & error budget, scripting, backups & DR. Soft: calm in incidents, blameless culture, automation mindset, crisis communication, discipline.
Artifacts
CI/CD pipelines, IaC manifests, dashboards & alerts, runbooks, postmortems. Pairs with the Release Manager and Backend; prod-readiness is part of DoD.
How AI / vibe-coding boosts the role
Pipeline generation; IaC draft; alert/log triage; runbook/postmortem; optimization — with ready prompts.
Growth: Junior → Middle → Senior → Lead
Junior: maintains pipelines & infra by example. Middle: builds CI/CD & IaC, monitoring, leads incidents. Senior: designs infra & reliability (SLO/error budget). Lead/Head of Platform: platform & reliability company-wide.
Common mistakes
Manual deploys & snowflake servers; no monitoring; postmortem as blame; untested recovery; irreversible ops without a safety net.
What to learn
CI/CD, IaC, containers & orchestration, observability, SLO/SLI/error budget, incident management, backups/DR, 12-factor. Read: Google SRE Book; The Phoenix Project; The DevOps Handbook; Accelerate.
Salary (RU)
Junior ~120–180k₽/mo, Middle ~180–320k, Senior ~320–500k+. One of the higher-paid roles; varies — check current data.
Laskoff agent mapping
No direct mapsTo; ship reliability is the house rule "backup/branch before anything irreversible" + the deterministic guard-dangerous.sh hook that blocks a catastrophic command before it runs.
🤖 Persona prompt
You are an experienced DevOps/SRE engineer. Help me make prod stable and releases fast through automation, not manual toil. Describe everything as code (IaC, pipelines); no snowflake servers. Bake in observability (metrics, logs, alerts) and SLOs before prod. Run incidents blameless: root cause (5 whys) and action items, not blame. Propose backups together with recovery verification. Before irreversible ops, demand a safety net (backup/branch). On request give pipelines, IaC, runbooks, postmortems and optimization tips.