AI Research Engineer

Smart Bricks; Dubai; On-site; AED 661,000 - 992,000; Full time

Job summary

Smart Bricks in Dubai is building an applied AI platform for real estate, combining ML research with production systems to advance domain-specific reasoning and financial analysis. You will work between research and production, developing evaluation suites, experimenting with frontier models, and turning research into reliable, scalable production capabilities for real estate decision making.

You will collaborate with engineers and domain experts to drive impact and iterate quickly based on

Qualifications

• Strong ML and deep learning foundations.
• Experience with LLMs, post-training, evaluations or applied NLP.
• Strong Python and modern ML frameworks.
• Experimental rigour and quantitative thinking.
• Ability to translate research into production impact.

Responsibilities

• Develop evaluation suites for real estate and financial reasoning.
• Experiment with frontier and open-source models.
• Fine-tune and adapt models for domain-specific tasks.
• Research reasoning, planning, memory and agent architectures.
• Build synthetic and expert-labelled datasets.
• Develop techniques for increasing reliability and reducing hallucination.
• Benchmark models across cost, latency and performance.
• Turn successful research into production systems.

Skills

ML foundations

Deep learning

LLMs / NLP

Python

Experimental rigour

Production impact

Tools

PyTorch

TensorFlow

JAX

Job description

Smart Bricks is the applied AI lab for real estate.

We are building the intelligence infrastructure for the world’s largest asset class. Our platform combines four intelligences: Company Brain, Real Estate Intelligence, Market Intelligence and Specialist Agents, with Financial Engines under the agents.

We operate globally across Dubai, London and San Francisco, bringing together exceptional people from AI, engineering, real estate and finance.

We care more about exceptional ability, rate of learning and ownership than conventional credentials.

The role

Your job is to push the performance frontier of Smart Bricks’ intelligence systems.

You’ll work between research and production, exploring how models can become materially better at real estate reasoning, financial analysis, document understanding and long-horizon agentic tasks.

What you’ll do

• Develop evaluation suites for real estate and financial reasoning
• Experiment with frontier and open-source models
• Fine-tune and adapt models for domain-specific tasks
• Research reasoning, planning, memory and agent architectures
• Build synthetic and expert-labelled datasets
• Develop techniques for increasing reliability and reducing hallucination
• Benchmark models across cost, latency and performance
• Turn successful research into production systems

What we’re looking for

• Strong ML and deep learning foundations
• Experience with LLMs, post-training, evaluations or applied NLP
• Strong Python and modern ML frameworks
• Experimental rigour and quantitative thinking
• Ability to translate research into production impact

Publications are welcome. They are not required.