Sr Software Development Engineer - AgentCore, AWS Agentic AI

About the role

Amazon Bedrock AgentCore is building the foundational services developers use to create, deploy, and operate production-ready AI agents at scale. AgentCore harness, launched at AWS Summit New York in June 2026, provides a managed agent loop that enables customers to define an agent through configuration instead of building and operating orchestration infrastructure themselves. Customers specify their model, instructions, tools, skills, memory, and execution limits. AgentCore harness manages the agent loop, secure execution environment, session isolation, identity, networking, observability, and deployment lifecycle. This allows developers to move quickly from an idea to a production agent while retaining the flexibility to integrate custom tools, models, and environments. As a Senior Software Development Engineer, you will design and build core components across the harness control plane and data plane. You will solve complex distributed systems problems involving agent execution, session state, tool orchestration, security, scalability, reliability, and low-latency streaming. You will help define how developers build and operate agents on AWS and influence the technical direction of a rapidly evolving product area. Beyond building harness itself, you will influence the technical direction and developer experience of the broader AgentCore organization. Harness sits at the intersection of AgentCore Runtime, Memory, Gateway, Identity, Observability, Browser, Code Interpreter, and Evaluations. You will work across these services to establish consistent abstractions, shape cross-service APIs, and remove friction from the complete agent development lifecycle. A central part of this role is making sophisticated agent capabilities dramatically easier to use. You will work backward from customer needs to turn complex infrastructure and orchestration tasks into intuitive configuration, sensible defaults, reusable templates, and guided workflows. Your work will help customers move from an idea to a secure, observable, production-ready agent in minutes, while preserving the flexibility advanced builders need for customization. You will identify the next set of capabilities that can make agent development faster and more accessible. This may include simplifying how customers connect tools and memory, introducing richer agent templates and skills, automating permissions and resource wiring, shortening the build-test-deploy loop, and creating safer ways to evaluate and release changes. This is a rare opportunity to join a newly launched service at an early stage, shape its architecture and customer experience, and deliver capabilities with broad industry impact. Machine learning experience is helpful but not required. We are looking for strong engineers with deep distributed systems expertise, sound technical judgment, and the ability to deliver in a fast-moving environment. Key job responsibilities - Design and build highly available control plane APIs, data plane services, and managed agent loops for production AI agents. - Deliver secure, isolated execution with tool orchestration, memory, context management, streaming, identity, networking, observability, and failure handling. - Integrate Amazon Bedrock and third-party models with AgentCore services, MCP servers, AWS Step Functions, Strands Agents, and customer-defined environments. - Design versioning, endpoints, progressive deployment, rollback, evaluations, and optimization capabilities that help customers safely improve agent behavior. - Improve latency, scalability, availability, resource efficiency, and cost controls across multi-tenant workloads. - Simplify the developer journey across APIs, SDKs, CLI, console, infrastructure as code, and local workflows through cohesive abstractions, templates, defaults, and automation. - Work directly with customers to reduce friction, accelerate time to production, and balance beginner-friendly experiences with flexibility for advanced builders. - Influence AgentCore-wide architecture and product strategy while raising the engineering bar through design reviews, operational readiness, code reviews, and mentoring.
- 5+ years of non-internship professional software development experience - 5+ years of programming with at least one software programming language experience - 5+ years of leading design or architecture (design patterns, reliability and scaling) of new and existing systems experience - Experience as a mentor, tech lead or leading an engineering team