Senior Software Engineer - Reliability

About this job

As a Senior Software Engineer - Reliability at Atlan in India, you will work on a multi-agent AI SRE platform that handles incident investigation, remediation, and networking tickets. This role is not an incident-command position where you manually fix production, nor is it a greenfield project; you will be enhancing a mature, opinionated codebase. On a normal day, your work involves improving root-cause diagnosis accuracy so engineers can trust agent investigations without double-checking, expanding auto-remediation playbooks safely with staged approvals, building fault-injection benchmarks and eval harnesses, and designing production guardrails like fail-closed checks and blast-radius limits. You will also develop new agents to eliminate operational toil and provide clean, reliable platform interfaces for other teams at Atlan.

Responsibilities

• Push root-cause accuracy in investigation agents so engineers trust the output without re-checking.
• Grow auto-remediation coverage safely through dry runs and human-approved stages.
• Build and run fault-injection benchmarks, eval harnesses, and validation mechanisms.
• Design production safety gates including fail-closed checks, kill switches, and blast-radius limits.
• Build new agents that target categories of operational toil and reduce future build costs.
• Create clean interfaces and safe defaults so other teams can adopt the reliability platform.
• Address urgent production issues by stopping the bleeding and removing the underlying failure class.

Requirements

• Experience living operational toil, carrying pagers, or owning incidents end to end.
• Track record of removing entire classes of recurring operational problems rather than speeding up runbooks.
• Ability to measure success by adoption metrics and real user impact with concrete numbers.
• Structural integration of AI into your own work, including shipping AI-native workflows or agentic guardrails.
• Failure-mode humility with a strong grasp of false-positive rates, rollback paths, and blast radiuses.
• Experience or instinct for building platforms and safe defaults for other teams' production systems.
• Ability to raise the bar and contribute quickly within an existing, high-rigor codebase.

Good to have

• Prior experience getting other engineering teams to adopt a built platform
• Direct background in architecting autonomous agents that take action with proper guardrails

Who should apply

This role suits engineers who have lived operational toil and want to make entire categories of production problems disappear using agentic systems. You should have a strong grasp of failure modes, blast radiuses, and AI guardrails, as you will be building a production-grade multi-agent SRE platform. This position is not for those who seek manual incident-command duties, prefer designing without shipping, or have experience with agents but have never owned the consequences of an agent failing in production. The posting does not mention whether freshers can apply.

Frequently asked questions