Imagine running search and analytics at any scale without thinking about clusters, capacity, or version upgrades. Where your queries return in milliseconds whether you're indexing a thousand documents or a hundred billion. Where the platform scales, heals, and secures itself so your team can focus on the questions, not the infrastructure.
That's what we build.
Amazon OpenSearch Service is the fully managed and Serverless AWS service that lets customers deploy, operate, and scale OpenSearch for log analytics, full-text search, application monitoring, observability, and AI-powered retrieval. We serve hundreds of thousands of customers running mission-critical workloads across every AWS region, and we are part of the AWS Database and Analytics organization.
Come join the OpenSearch Control Plane team. We are responsible for a service that:
- Reliably manages a large fleet of cloud-native OpenSearch clusters and collections, freeing customers from sizing, scaling, and patching decisions
- Guarantees high availability and durability for mission-critical search and observability workloads at the scale of the most demanding internet businesses
- Orchestrates and automates the complete lifecycle of an OpenSearch cluster or collection — from creation through scale-up, scale-out, upgrade, replication, and fail-over
The OpenSearch control plane is not just any distributed system. It orchestrates a fleet of clusters across every AWS region, detects and recovers from node failures in seconds, and serves workloads that span petabytes of customer data. Most recently, we launched **OpenSearch Serverless NextGen** — delivering ultra-fast collection provisioning and true scale-to-zero capacity, so customers pay only for what they use and get a working environment in seconds rather than minutes. We are not done; this is one of many investments raising the bar for what customers can expect from a managed search service.
OpenSearch itself is a community-driven, Apache 2.0–licensed open-source search and analytics suite built on Apache Lucene. Since launching in July 2021, the project has delivered multiple major and minor releases advancing core search, vector search, and analytics. Our next phase focuses on transforming OpenSearch into an intelligent retrieval engine — one that meets the evolving needs of customers running search, analytics, and AI-powered workloads across structured, unstructured, and multi-modal data.
We work in Java, Rust, Python, and Go. We contribute upstream to OpenSearch and Apache Lucene. We are customer-obsessed, operate with the entrepreneurial pace of a startup inside one of the world's largest cloud providers, and value strong intuition backed by metrics.
**Operations is core engineering on this team, not overhead.** Engineers operate what they build, and we treat customer-impacting incidents with the same engineering rigor we apply to feature work. Reliability and Static Stability is a forcing function for design, not a phase after launch.
Key job responsibilities
- Design, develop, and operate components of a large-scale distributed control plane — reason about failure modes, multi-tenant isolation, and blast-radius before writing code
- Look for opportunities to simplify — consolidate duplicate abstractions, remove dead code, reduce configuration sprawl, and challenge complexity that doesn't earn its keep
- Treat operations as core engineering — own the systems you build end-to-end, participate fully in incident response and RCAs, and drive operational improvements that reduce cost and maintenance burden
- Solve real problems in cluster orchestration, capacity provisioning, lifecycle workflows, multi-tenant infrastructure, query routing, and operational tooling
- Use AI coding assistants and GenAI tools effectively across the software development lifecycle — from design and prototyping through coding, testing, code review, debugging, and operations — applying engineering judgment to validate AI-generated output against Amazon's correctness, security, and operational bar
- Write robust, efficient, and maintainable code in Java, Rust, Python, or Go
- Partner with senior engineers to translate ambiguous requirements into concrete deliverables; prototype, test, and validate
- Continually challenge what exists and explore what should change to better serve customers