Software Engineer - Agent Development

Apple

1d

About The Position

Apple’s IS&T Customer Systems group is reimagining how Apple engineers work through intelligent, autonomous systems. The Software Engineer will design, build, and operate agentic AI systems end-to-end across client-side and server-side applications. The role focuses on architecting agents that reason, plan, remember, and act reliably across long-running workflows, with particular attention to memory design, multi-agent orchestration, observability, and maintainability.

This is a technical engineering role with a high bar for production quality. The engineer will own the lifecycle of agentic systems, from architecture and memory design through tool integration, observability, and safe deployment. The role also requires a solid understanding of Java, React, and TypeScript, and involves collaboration with engineers, product managers, and platform teams across Apple to bring AI capabilities into internal developer tools.

Requirements

• At least 4 years of software engineering experience, including at least 2 years focused on AI/ML systems, LLM integrations, or autonomous agent development.
• Hands-on experience designing and implementing multi-agent orchestration systems and agent loop patterns such as ReAct or Plan-and-Execute.
• Strong proficiency in Java (Spring Boot) and TypeScript/React for full-stack agentic applications.
• Understanding of LLM fundamentals, including prompt engineering, context-window management, function calling, structured outputs, and tool use.
• Experience building RAG pipelines and working with vector databases.
• Demonstrated ability to implement agent safety guardrails, including prompt-injection defences and output schema enforcement.
• Experience with observability tooling such as OpenTelemetry applied to distributed, AI-driven systems.
• Bachelor’s degree in Computer Science, Software Engineering, or a related technical field.

Nice To Haves

• Experience implementing or consuming MCP (Model Context Protocol) servers and clients.
• Familiarity with model routing and fallback strategies across multiple LLM providers.
• Reactive programming in Java, including Project Reactor and WebFlux, for streaming LLM responses.
• Python proficiency for agent prototyping and evaluation-pipeline scripting.
• Hands-on experience with cloud AI platforms.
• Experience with agent sandboxing and safe code-execution environments.
• Awareness of fine-tuning, LoRA, and small-model distillation techniques for agent specialisation.

Responsibilities

• Design and implement multi-agent orchestration, including collaboration, task delegation, hand-offs, shared context management, role specialisation, and conflict resolution.
• Build episodic, semantic, and working-memory architectures using vector stores such as pgvector and Milvus, and hybrid retrieval strategies.
• Develop tool and function-calling layers that expose internal APIs and services as agent-executable tools through MCP servers or custom tool registries, with clear contracts and failure semantics.
• Implement RAG pipelines using vector stores and hybrid retrieval to ground agent responses in authoritative internal knowledge.
• Own observability for agentic systems, including distributed tracing, prompt and response capture, token and cost accounting, evaluation pipelines, replay and debugging, and regression detection.
• Improve maintainability through versioning prompts, tools, and agent graphs; deterministic test harnesses; evaluation suites; safe canary rollouts; and structured logging.
• Apply safety guardrails, including input validation, prompt-injection defences, output schema contracts, and sandboxed code execution.
• Architect and maintain server-side agent services using Java, Spring Boot, and Spring AI, and client-side agent interfaces using React and TypeScript.
• Collaborate with engineers, product managers, and platform teams across Apple to deliver AI capabilities in developer tools.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level