AI Engineer

Job Title: AI Engineer (5+ Years' Experience)

Location

Dubai Investment Park, Dubai, UAE Experience Required: 5+ years Job Type: Full-Time, office-based Reports To: Group Manager Data Science & Governance Direct Reports: None Job Summary Amana is seeking an AI Engineer to design, build, and run the group's AI applications and predictive models on Microsoft Azure. You will take use cases from business problem to production generative AI and agent solutions on Microsoft Foundry, data and AI applications on Databricks Apps, and machine-learning predictions served from the group's Databricks lakehouse and keep them accurate, secure, and cost-effective once they are live. Through design, evaluation, and mentoring, you will set the quality bar for AI delivery in the team. This role sits within Amana's Data Science, Governance & AI Centre of Excellence, with a clear path into AI engineering leadership. Why This Role Matters Amana operates across capital-intensive sectors where cost, schedule, safety, and cash flow decide project outcomes. AI creates value only when it reaches the people who run projects, sites, and finance as tools they use every day, built on governed data. Your applications and models will put predictions and AI assistance into the hands of project, commercial, and operations teams across the group. You will turn governed data into working, trusted AI products and your designs and evaluations will shape how the rest of the team builds AI. Core Responsibilities AI Application Design & Delivery Design, build, and deploy generative AI and agent solutions on Microsoft Foundry model selection, prompt and tool design, retrieval-augmented generation (RAG) over governed Amana content, and Foundry Agent Service. Build and run data and AI applications on Databricks Apps interactive tools, RAG chat assistants, data-entry forms, and operational interfaces on lakehouse data governed by Unity Catalog. Deliver construction use cases such as contract review, document intelligence, and project risk prediction, integrated with Microsoft 365 and Power Platform so they reach users in the tools they already use. Own each solution end to end: requirements, design, build, user acceptance, and handover to support. Machine Learning & Predictive Analytics Develop, train, validate, and deploy machine-learning models forecasting, classification, and anomaly detection on Databricks using MLflow, Unity Catalog feature tables, and Model Serving. Run batch and real-time predictions in production and surface them where decisions are made: Power BI, Databricks Apps, and business workflows. Apply predictive analytics across project controls, cost and commercial, procurement, HSE, plant, and workforce data, and explain the method, confidence, and limitations of each model to business users. Evaluation, Responsible AI & Security Evaluate every AI solution before release for accuracy, groundedness, bias, and robustness, using Foundry evaluations and MLflow, and record the results. Monitor deployed models and agents for drift, failure, and answer quality, and retrain or roll back when agreed thresholds are breached. Build security and data protection in from design: Microsoft Entra ID and managed identities, least-privilege access, secrets management, content safety filters, and PDPL controls for personal and HR data. Partner with the Data Governance Office to take each new AI use case through Amana's data risk assessment, apply data-residency requirements to every model deployment, and record each solution in the AI use-case register. MLOps, Reliability & Cost Apply CI/CD, automated testing, and version control to code, prompts, and models using Git and Azure DevOps. Own the operational health and running cost of your solutions: tracing, monitoring, alerting, incident response, and inference and compute spend within agreed budgets. Document every solution architecture, data sources, prompts, evaluation results, and runbooks so it can be supported and handed over without loss of knowledge. Collaboration & Mentoring Work with business owners in finance, projects, procurement, HSE, and HR to identify high-value use cases and agree success measures before build. Partner with the Data Engineering Lead (Data Architect) and Senior Data Engineer on AI-ready data feature tables and document and vector pipelines and with the Data Analyst on Power BI integration. Mentor Data Scientists and engineers in AI engineering practice, and help colleagues adopt approved AI tools safely through the Amana Digital Academy. Explain design choices, trade-offs, risks, and model limitations clearly to technical and business audiences. Preferred Qualifications Hands-on experience with Databricks Apps, Mosaic AI Vector Search, Model Serving, and Unity Catalog. Agent orchestration and tool integration, including multi-agent workflows and Model Context Protocol (MCP) servers. Document intelligence on construction documents contracts, drawings, and specifications combining OCR with LLM-based extraction. Copilot Studio, Power Apps, and Power Automate for extending AI solutions into Microsoft 365. Time-series forecasting for project cost, schedule, or resource planning. Multi-entity, multi-country group exposure with mixed data residency considerations. Industry experience in construction, engineering services, manufacturing, energy, or logistics. Desired Candidate Profile 5+ years of hands-on experience in machine learning, AI, or software engineering, including at least 2 years building and deploying ML or generative AI solutions in production. Built and shipped generative AI or agent applications on Microsoft Foundry, Azure OpenAI, or an equivalent platform, including RAG, tool calling, and systematic evaluation. Developed and deployed machine-learning models end to end feature engineering, training, validation, deployment, and monitoring ideally on Databricks with MLflow. Strong Python and SQL, with experience building web applications or APIs (for example Streamlit, Dash, FastAPI, or React) and Git-based CI/CD delivery (Azure DevOps or equivalent). Security and responsible AI applied in practice: Microsoft Entra ID and managed identities, secrets management, content safety, and testing for groundedness and bias. Excellent written and verbal communication skills across technical and business audiences.