Backend Engineer, Memory Integration

Every agentic AI platform faces the challenge of making memory work at scale without sacrificing security or speed. The Backend Engineer, Memory Integration role in Portugal is remote and sits at the intersection of agent framework design and production-grade AI systems. This is where Python libraries and memory-enabled agent templates shape how context and session data are handled across the platform. The person in this role shapes how agent memory becomes a seamless, reliable part of every agent’s workflow.

Responsibilities

• Integrate agent memory with write-on-turn (store session events) and retrieve-on-turn (fetch relevant context)
• Create, query, update and delete memory using AWS AgentCore Memory API
• Design memory namespaces for per-agent isolation and per-user scoping
• Develop Python helpers for agent memory in the platform library
• Integrate memory into agent templates to enable configuration-based memory without custom code

Requirements

• Hands-on experience with agent memory integration in a production agentic AI project, such as AWS AgentCore Memory API (write-on-turn/retrieve-on-turn), memory namespace design or agent template memory configuration (general backend/API integration experience without agent-specific memory context does not meet this bar)
• Three-plus years of backend engineering with AI or machine learning systems
• API integration for stateful services
• Session management and context persistence
• Python software development kit (SDK) development

Nice to have

• AWS AgentCore Memory SDK usage
• LangChain memory integration
• Agent framework development (LangGraph, Crew)
• Conversational AI context handling