AI Engineer

Position Overview

We build AI Agents on an internal Agentic AI platform. The platform supports creating and configuring agents, connecting capabilities via MCP, and enriching agent responses using a file-based knowledge base with built-in document and data RAG capabilities. We focus on building high business value, production-ready agents, integrations, and evaluation workflows.

We are looking for a strong mid-level AI Engineer with a hybrid profile — someone who can both build production-ready AI solutions and lead business-facing conversations to identify, qualify, and shape high-value use cases. You will design and build AI Agents on our internal Agentic AI platform, develop MCP servers, engineer prompts and agentic workflows, and work directly with business stakeholders to extract requirements, guide business case development, and assess the strategic value of proposed AI use cases.

Key Tasks and Responsibilities

Business Engagement & Use Case Development

• Lead structured discovery sessions with business stakeholders to extract requirements and map existing processes
• Assess and qualify inbound AI use case requests, evaluating feasibility, complexity, and business value
• Guide business teams through the process of building a business case for agentic AI initiatives
• Facilitate workshops and conversations to translate business problems into well-defined AI solution briefs
• Act as the bridge between business and engineering, communicating technical constraints and possibilities in accessible business language
• Collaborate with business stakeholders to gather requirements and translate business processes into agent workflows
• Design agent conversation flows and user experiences that align with business objectives, and iterate based on user feedback and business success metrics

AI Engineering

• Create and configure new AI Agents on our Agentic AI platform
• Design, test, and iterate agent instructions/prompts (context engineering) to ensure correct and consistent behaviour
• Connect and configure tools via MCP, including defining tool contracts, validating behaviour, and troubleshooting failures
• Build new MCP servers when needed by wrapping APIs/flows into tools with proper authentication, error handling, and logging
• Design end-to-end agentic workflows incorporating analytics pipelines and AI models
• Upload, organise, and maintain files in an agent knowledge base to improve grounding and retrieval quality
• Create and maintain search indexes for file corpora when required, including manual index setup and refresh processes
• Set up and run evaluations using MLflow, including offline test sets, regression checks, and metrics tracking

Requirements

Business & Communication Skills

• Strong communication skills with the ability to lead business conversations, run discovery workshops, and translate requirements into technical specifications
• Experience in business analysis, requirements gathering, or customer-facing technical roles
• Ability to assess the business value and feasibility of AI use cases, including ROI framing and prioritisation
• Ability to think from the end-user perspective and optimise agent UX for business outcomes
• Comfortable facilitating workshops and leading conversations with non-technical stakeholders

AI Engineering

• 3+ years of professional software engineering experience, with a backend focus; Python preferred
• 2+ years building LLM-powered applications, including AI agents with tool use and RAG/knowledge base grounding
• Solid understanding of LLMs and their behaviour, including tokens/context windows, prompting patterns, hallucinations and grounding, and latency/cost trade-offs
• Strong experience with prompt/instruction design and iteration (context engineering) and debugging agent behaviour
• Solid understanding of agent architectures, including planning, tool execution, memory/context, retrieval grounding, and safety/guardrails, with hands-on delivery experience
• Hands-on experience integrating external systems and tools via MCP, including connecting tools and designing/implementing MCP servers
• Experience implementing secure integrations using OAuth 2.0 / OIDC and understanding identity concepts such as Azure Entra ID, including apps, tenants, scopes/roles, and consent
• Solid software engineering fundamentals, including Git, code reviews, testing, logging/monitoring, and performance-aware development
• Good spoken English and ability to work in a highly integrated Agile team

Nice to Have

• Experience with one cloud provider, preferably Azure, AWS, or GCP, for deploying or operating services
• Experience using an agentic framework such as LangChain, Semantic Kernel, Autogen, CrewAI, or similar
• Experience with search indexing/retrieval systems, including schema design, ingestion pipelines, and index refresh automation
• Familiarity with containerisation (Docker) and basic CI/CD practices for backend services
• Exposure to AI value frameworks or benefit realisation methodologies
• Experience working in a Centre of Excellence (CoE), innovation, or transformation function
• Familiarity with AI governance considerations when assessing use cases, including data privacy, risk, and compliance
• Experience with MCP server development from scratch, not just configuration

What CRH Offers You

• A culture that values opportunity for growth, development, and internal promotion
• Comprehensive secondary benefits
• Significant contribution to your pension plan
• Health and wellness programs, including an on-site gym and fitness classes
• Excellent opportunities to develop and progress with a global organization

Benefits/perks listed above may vary depending on the nature of the employment with CRH and the country where you work.