Machine Learning Engineer

Client of Involved Solutions; Dubai; On-site; AED 180,000 - 300,000; Full time

Job summary

Client of Involved Solutions is partnering with a fast-scaling UAE technology business to hire a Machine Learning Engineer. The role centers on cutting-edge applied AI, deploying models to live production and measuring commercial impact.

You will fine-tune foundation models, build agentic systems with tool use and multi-step reasoning, and optimise inference on GPU infrastructure with autonomy and strong compute access.

Qualifications

• 5 years of hands-on experience in machine learning engineering, with a track record of deploying models into production.
• Strong Python and deep experience with PyTorch, plus Hugging Face Transformers or equivalent.
• Practical experience with LLMs in production, including fine-tuning, RAG and evaluation.
• Solid grounding in MLOps and cloud ML platforms on AWS, Azure or GCP.
• Experience with model serving and inference optimisation, or a clear appetite to go deep on it.
• Exposure to multimodal models, distributed training or GPU cluster work is a strong advantage.
• Based in the UAE or willing to relocate.

Responsibilities

• Fine-tune and adapt foundation models using LoRA, QLoRA, DPO and RLHF.
• Build and ship production agentic systems with tool use and multi-step reasoning.
• Optimize inference for latency and cost with quantisation and distillation.
• Design evaluation frameworks for model quality, safety and drift.
• Develop the MLOps backbone: training pipelines, CI/CD for ML and monitoring on GPU infrastructure.
• Collaborate with product and engineering to turn business problems into ML solutions.

Skills

Python

PyTorch

Transformers

LLMs

MLOps

AWS

Azure

GCP

Model serving

GPU clusters

Multimodal

Distributed training

vLLM

TensorRT-LLM

Tools

LoRA

QLoRA

DPO

RLHF

vLLM

TensorRT-LLM

Job description

We have partnered with a fast-scaling technology business in the UAE to hire a Machine Learning Engineer. This is a team working on the hard end of applied AI: fine-tuning foundation models, building agentic systems that take real actions, and pushing inference to run faster and cheaper on serious GPU infrastructure. The models you build will be live in production, used by real customers and measured on commercial outcomes.

This role suits an engineer who wants to work with the current generation of ML tooling, not maintain a legacy pipeline. You will have the compute, the data and the autonomy to do it properly.

About the role:

• Fine-tune and adapt open-source and proprietary foundation models using techniques such as LoRA, QLoRA, DPO and RLHF, including multilingual and Arabic language use cases
• Build and ship production agentic systems with tool use, multi-step reasoning, retrieval and orchestration
• Optimise inference for latency and cost using quantisation, distillation and serving frameworks such as vLLM and TensorRT-LLM
• Design evaluation frameworks that measure model quality, safety and drift in live environments
• Build the MLOps backbone covering training pipelines, experiment tracking, model registries, CI/CD for ML and monitoring on GPU infrastructure
• Work with product and engineering teams to turn business problems into ML systems that ship and move the numbers

Desired Candidate Profile

About you:

• 5 years of hands-on experience in machine learning engineering, with a track record of deploying models into production
• Strong Python and deep experience with PyTorch, plus Hugging Face Transformers or equivalent
• Practical experience with LLMs in production, including fine-tuning, RAG and evaluation
• Solid grounding in MLOps and cloud ML platforms on AWS, Azure or GCP
• Experience with model serving and inference optimisation, or a clear appetite to go deep on it
• Exposure to multimodal models, distributed training or GPU cluster work is a strong advantage
• Based in the UAE or willing to relocate