AI Engineer in Abu Dhabi at InfinitePL

ABOUT INFINITE PL
Infinite pl is a digital-led tech firm driven to become a digital logistics pioneer by harnessing the power of people, data, and platforms. We orchestrate and build innovative platforms that tackle complex problems within logistics and adjacent sectors, enriching the experiences of governments, businesses, and residents through cutting-edge digital solutions.

ROLE SUMMARY
We are looking for an AI Engineer who can turn priority use cases into working, evaluable agents. You will design context, tools, prompts, retrieval flows, agent loops, and evaluation bundles, and contribute to reusable agentic patterns that let us scale safely across Fast Track innovation and Production-grade delivery. This is a hands-on builder role for someone who pairs strong software engineering with practical LLM and agentic AI delivery.

• Design and build agentic AI solutions from concept to pilot or production, including agent role definition, autonomy boundaries, tool/data scopes, guardrails, and evaluation criteria.
• Implement agent loops, tool-use patterns, context engineering, retrieval/RAG flows, HITL approval journeys, tracing, telemetry, and audit capture.
• Select and justify models based on latency, cost, residency, reliability, and use-case fit; tune solutions for performance and operational cost.
• Develop reusable Agent JD templates, prompt/context patterns, orchestration patterns, and evaluation assets that can be adopted across teams.
• Partner with governance, platform, and integration teams to make agents gate-ready by design, including prompt-injection, PII leakage, and reliability controls.
• Support complex and multi-agent solutions, including orchestration, memory, retrieval, reasoning, and agent-to-agent composition.
• Move solutions through sandbox, staging, and production promotion paths, ensuring each agent is observable, versioned, tested, and auditable.
• Mentor junior engineers and contribute to a strong engineering culture around quality, speed, learning, and responsible AI delivery.

• Strong software engineering background in Python and/or C#, with experience building production-grade APIs, services, or cloud-native applications.
• Hands-on experience with LLMs, agent frameworks, and agentic design patterns such as Semantic Kernel, Microsoft Agent Framework / AutoGen, LangGraph, or comparable frameworks.
• Practical experience with RAG, vector search, prompt/context engineering, system prompt design, tool calling, and model evaluation.
• Ability to build and maintain evaluation sets, regression tests, and acceptance criteria for agent behaviour, reliability, and safety.
• Experience integrating APIs, data sources, and tools into AI workflows; comfortable debugging across application, data, and model layers.
• Understanding of cloud-native delivery, observability, CI/CD, secrets management, telemetry, and secure software development practices.
• Strong communication skills with the ability to explain technical trade-offs to product, business, governance, and leadership stakeholders.

• Azure AI Foundry, Azure OpenAI, Azure AI Search, Functions, Container Apps, Cosmos DB, Redis, OpenTelemetry, or equivalent cloud AI stack experience.
• Experience with bilingual or Arabic/English AI products, evaluation design, and user-facing AI experiences.
• Experience with multi-modal AI, including vision, voice, document intelligence, or video/avatar experimentation.
• Prior work in government, regulated, sovereign cloud, or enterprise environments where auditability and data residency matter.
• Technical leadership or player-coach experience, including mentoring engineers and raising engineering standards.

Azure AI Foundry - Semantic Kernel - LangGraph - Azure OpenAI - AI Search - Cosmos DB - Redis - Functions, Container Apps - Eval harness - OpenTelemetry

• At least one agent progresses through the full factory path into production or production-equivalent validation on a real source, with a passing evaluation bundle.
• A reusable Agent JD, prompt/context, or orchestration pattern is contributed to the team library.
• The AI engineering team improves first-time gate readiness through better patterns, testing, and documentation.

This role is not. A pure model research role or a platform/landing-zone owner. This role builds working agents and the patterns that make them reliable.