AI/ML Agent Engineer — Back-End Focus at Ebay in Dublin

Job description

• We’re hiring a Junior AI/ML Agent Engineer with a back-end focus to build reliable agent services and APIs using Java Spring Boot. You’ll implement agentic workflows powered by LLMs, OpenAI specs, and MCP, integrating with developer tools like GitHub, Cline, and Claude models.

Responsibilities

• Develop Spring Boot microservices that orchestrate LLM-powered agents using OpenAI specs and MCP (clients/servers, tool schemas).
• Implement tool execution layers, function calling, retries, guardrails, and safety checks for robust agent behavior.
• Build APIs and event-driven pipelines for agent requests, context assembly (RAG/embeddings), and result streaming.
• Integrate with GitHub (webhooks, Actions, PR review bots), Cline workflows, and Claude endpoints for developer-focused automations.
• Manage configuration, secrets, and access control; ensure compliance with internal security standards and least-privilege design.
• Add observability: logs, metrics, traces, and run metadata; create dashboards for reliability and performance.
• Write comprehensive tests (unit/integration), automate builds/deployments, and maintain documentation for specs and service contracts.
• Collaborate with front-end and platform teams to deliver end-to-end features; participate in design and code reviews.

Requirements

• 2–5 years of professional experience in back-end engineering with Java and Spring Boot.
• Practical experience integrating LLMs (OpenAI, Claude) into services; understanding of agentic workflows and tool/function orchestration.
• Familiarity with MCP and implementation of clients/servers, plus OpenAI spec-driven tool definitions.
• Experience with REST/GraphQL APIs, data modeling, and message/event systems (e.g., Kafka/SQS).
• Exposure to GitHub APIs/Actions, Cline, and Claude model integrations.
• Solid grasp of testing, CI/CD, and cloud deployment patterns.
• Experience with RAG, embeddings, or lightweight vector stores; caching and rate-limiting for cost/perf control.
• Knowledge of security for AI services: prompt injection defenses, input validation, and secret management.
• Observability with Datadog/OpenTelemetry or similar; performance tuning in JVM/Spring.

Skills

Required

• Java
• Spring Boot
• LLM integration (OpenAI, Claude)
• Agentic workflows
• Tool/function orchestration
• MCP (clients/servers)
• OpenAI spec-driven tool definitions
• REST/GraphQL APIs
• Data modeling
• Message/event systems (e.g., Kafka/SQS)
• GitHub APIs/Actions
• Cline

Preferred

• RAG
• Embeddings
• Lightweight vector stores
• Caching
• Rate-limiting
• Observability (Datadog/OpenTelemetry)
• Performance tuning in JVM/Spring

Benefits

• Competitive junior-level compensation with benefits. Final level and pay depend on experience and interview performance.