AI Engineer

Key Responsibilities

Agentic AI Design & Development

• Design and implement autonomous AI agents capable of reasoning, planning, and executing goal-driven actions with minimal human intervention
• Develop multi-step workflows combining LLM reasoning, tool use, and structured decision-making
• Build and maintain prompt engineering strategies, including chain-of-thought, role-based prompting, and instruction design
• Implement memory architectures (short-term and long-term) for context-aware agent interactions
• Develop evaluation frameworks to measure agent performance, accuracy, and reliability

Platform Integration & RAG Systems

• Build retrieval-augmented generation (RAG) pipelines connecting agents to organisational knowledge bases and document repositories
• Integrate agents with internal APIs, data sources, and enterprise systems
• Design and optimise vector search strategies using embedding models and vector databases
• Implement document processing pipelines (ingestion, chunking, indexing, retrieval)
• Develop connectors for heterogeneous data sources (structured databases, unstructured documents, web content)

Multi-Agent Orchestration

• Design multi-agent architectures where specialized agents collaborate to accomplish complex tasks
• Implement agent coordination patterns (sequential, parallel, hierarchical, event-driven)
• Build routing and delegation logic for directing user requests to appropriate specialized agents
• Define inter-agent communication protocols and shared state management
• Optimise agent workflows for latency, cost, and quality trade-offs

Managed Agentic Platforms

• Leverage managed agentic services such as AWS Bedrock AgentCore and Microsoft Foundry Agent Service
• Evaluate and recommend platform capabilities for specific use cases (tool integration, knowledge bases, guardrails)
• Design hybrid architectures combining managed services with custom agent implementations
• Stay current with evolving managed platform features and incorporate relevant capabilities

Safety, Governance & Monitoring

• Implement content safety guardrails, input/output filtering, and policy enforcement
• Design human-in-the-loop approval mechanisms for sensitive or high-impact agent actions
• Build observability pipelines: trace agent decisions, log tool calls, monitor performance
• Define and track evaluation metrics for agent quality and safety
• Contribute to data protection impact assessments when introducing new AI capabilities
• Ensure compliance with GDPR, data residency requirements, and institutional policies

Research Unit & Corporate Service Support

• Engage with research units and corporate services to identify workflows suitable for AI agent automation
• Translate domain-specific requirements into agent specifications and implementation plans
• Conduct discovery workshops and prototype demonstrations to validate agent approaches
• Provide technical guidance and training to users adopting AI agent capabilities
• Document agent capabilities, limitations, and best practices for end users

Required Skills & Experience

Technical Expertise

• Strong proficiency in Python and async programming patterns
• Demonstrated experience building and deploying LLM-based agents in production environments
• Hands-on experience with agent orchestration frameworks (LangChain, LangGraph, CrewAI, Strands Agents, or equivalent)
• Solid understanding of RAG architectures: embeddings, vector databases, chunking strategies, reranking
• Experience with managed AI services on AWS (Bedrock incl. AgentCore) and/or Microsoft (Foundry incl. Agent Service)
• Understanding of LLM capabilities and limitations across model families
• Experience with API design, REST/GraphQL, and integration patterns
• Familiarity with containerised deployments and Infrastructure as Code
• Knowledge of observability tools and practices (logging, tracing, metrics)

Proven Track Record (in lieu of years of experience)

• Portfolio of shipped agent systems: production deployments, open-source projects, or documented enterprise implementations
• Demonstrated ability to take an agent from concept through design, development, and production deployment
• Evidence of solving real-world problems with agentic AI (GitHub repositories, case studies, technical blog posts, or conference presentations)
• Experience working autonomously on complex technical problems with minimal supervision

Nice to Have

• Experience with multi-cloud environments (AWS + Azure)
• Familiarity with Open WebUI, LiteLLM, or similar open-source AI platforms
• Understanding of European data protection regulations (GDPR)
• Contributions to open-source AI/ML projects

Certifications (Preferred, not required)

• AWS Certified Generative AI Developer - Professional
• Azure AI Engineer Associate
• Any relevant specialisation in generative AI or agentic systems

Soft Skills

• Strong problem-solving mindset with ability to navigate ambiguity
• Effective communication with both technical and non-technical stakeholders
• Ability to work autonomously and manage priorities in a fast-evolving field
• Collaborative approach to supporting diverse teams (researchers, engineers, corporate staff)
• Curiosity-driven: proactive in exploring new tools, frameworks, and approaches
• Fluency in French and English