Site Reliability Engineer

Location: Hyderabad, India; Indore, India

Department: Engineering

Experience: 2+

• Availability is your scoreboard. You define SLIs/SLOs, own the error budget, and make the call on when to slow down shipping to protect it.
• Deploys don't cause incidents. You build CI/CD that ships Blue-Green, Canary, and Rolling releases with zero customer-facing downtime — and you built at least one of these pipelines from scratch before, not just configured someone else's.
• Infrastructure is code, not tribal knowledge. Terraform and Ansible define your environments; nothing gets clicked into existence in a console.
• You see problems before customers do. Prometheus, Grafana, ELK, Datadog, OpenTelemetry — instrumented well enough that alerts mean something and noise doesn't.
• Incidents end with a fix, not a Slack thread. You lead response, drive RCA, and turn postmortems into action items that actually close.
• Cost is engineered, not just monitored. You right-size and forecast capacity instead of reacting to the bill.
• You use AI to move faster, not to look modern. Alert triage, incident summarization, automated runbooks — if it can be scripted or delegated to a model, it should be.
• On-call is shared, not survived. You rotate in, and you're the person newer engineers want on the call at 2am.

• 2+ years running production as an SRE, DevOps, Platform, or Cloud Engineer — real ownership, not observer status
• Production-grade AWS, Azure, or GCP experience
• Docker/Kubernetes under actual load, with the scars to prove it
• Terraform and Ansible (or equivalent IaC) in daily use
• Built CI/CD pipelines from zero — Jenkins, GitHub Actions, GitLab CI, or similar
• Solid Linux, networking, and distributed-systems fundamentals
• Python, Go, or Bash for scripting and automation
• Shipped Blue-Green, Canary, and Rolling deployments in production, not just in theory
• Domain background in FinTech, Payments, Healthcare, or another high-availability environment
• AI tools (Copilot, Claude, Cursor, ChatGPT) are already part of how you work, not a novelty

• You've shipped AI-powered ops tooling — triage bots, auto-summarized incidents, anomaly detection that actually fires correctly
• You talk SLOs, error budgets, and chaos engineering like it's your first language, because it is