Amazon OpenSearch Service powers search, log analytics, observability, and vector workloads for tens of thousands of customers. Every one of those customers runs on software that our Data Plane Platform team ships, and keeping it current means updating and managing software updates across millions of instances at scale, safely and continuously. We own the release engine for the managed OpenSearch data plane: the pipelines, the automation, and the self-service release platform that take an upstream open source cut all the way to GA across every supported version and region.
We are looking for a Software Development Engineer to help us turn releasing OpenSearch from a hand-coordinated effort into a fully automated, self-service platform. The team has already sharply reduced release cycle time, increased release throughput, consolidated its pipelines, and driven per-engineer release effort close to zero. You will build pieces of that system: the tooling, orchestration, and AI-assisted automation that let teams release safely with less manual effort.
This is a hands-on builder role at the intersection of distributed systems, CI/CD, and developer experience, with a strong and growing emphasis on AI. You will own well-scoped components end to end, from design through operations, with mentorship from senior engineers on the team, and you will have real opportunity to apply AI and agentic automation to real release problems. The code you ship becomes part of the paved road the OpenSearch organization releases on.
Key job responsibilities
Build and operate the services, pipelines, and automation that update and manage software updates across millions of instances at scale, safely and continuously, across many versions and regions.
Build and extend the self-service release platform so builders across the organization can ship features quickly and reliably.
Improve release safety with staged rollouts, automated validation, rollback, and blast-radius controls so a bad change does not reach customers at scale.
Apply AI and agentic automation to the parts of releasing that still require manual effort today.
Own the operations of what you build: on-call, runbooks, metrics, and reducing operational load over time.
Partner with the engineering teams who build and release on top of your platform.
