Agentic AI Engineer — LLM Systems in Production

mindX360; Dubai; On-site; AED 210,000 - 360,000; Full time

Job summary

Emaar Group is seeking a frontend-like AI/ML engineering role to design, build, and ship production-grade, LLM-powered agentic applications across the enterprise. You will drive multi-agent orchestration, tool-calling loops, and code-execution sandboxes, built on frameworks such as Claude Agent SDK, with a focus on reliable, secure deployments.

You will own end-to-end data paths, including ingestion, embeddings, vector stores, and evaluation, and partner with business stakeholders to deliver

Qualifications

• Strong Python and systems engineering background with production-grade software experience.
• Hands-on LLM and agent experience—orchestration, RAG, evals, and prompt engineering.
• Experience with data pipelines, embeddings, vector stores, and production AI operations.
• Comfort with containerization (Docker) and CI/CD-based deployment of AI workloads.
• Ability to work across ambiguities and communicate outcomes to non-technical stakeholders.

Responsibilities

• Design and ship LLM-powered agentic applications with multi-agent orchestration and tool-calling loops.
• Architect RAG pipelines over enterprise data, including retrieval, context, and guardrails.
• Own full data path: ingestion, embeddings, vector stores, evaluation.
• Engineer prompts, context strategies, and agent memory for robust orchestration.
• Build evaluation harnesses and production dashboards for insights.

Skills

Python

LLM engineering

Data pipelines

CI/CD

Docker

Multi-agent orchestration

Stakeholder communication

Education

Bachelor’s degree in CS/AI/Data Science/Engineering

Tools

Docker

Kubernetes

Claude Agent SDK

Job description

Emaar Group is seeking a frontend-like AI/ML engineering role to design, build, and ship production-grade, LLM-powered agentic applications across the enterprise. You will drive multi-agent orchestration, tool-calling loops, and code-execution sandboxes, built on frameworks such as Claude Agent SDK, with a focus on reliable, secure deployments.

You will own end-to-end data paths, including ingestion, embeddings, vector stores, and evaluation, and partner with business stakeholders to deliver