Lead Data Intelligence ML Engineer: Auto-Labeling Pipelines

Dyson; Dubai; On-site; AED 420,000 - 700,000; Full time

Job summary

Dyson seeks a Lead Data Intelligence ML Engineer to design automated data-labelling pipelines, reducing manual annotation via Active Learning and Weak Supervision. You will bridge data collection and model-ready datasets, ensuring high-quality labels at scale.

You will collaborate with Dyson’s global engineering teams and external partners, shaping data strategies, and building robust data prep and MLOps infrastructure in a fast-moving environment.

Qualifications

• 8+ years in ML engineering focused on data-centric AI or CV/NLP pipelines.
• Strong Python, ML stack expertise (PyTorch/TensorFlow, NumPy, Pandas, scikit-learn).
• Experience with automated labelling, weak supervision, or active learning.
• Data engineering with large-scale unstructured data; SQL/NoSQL.
• Cloud ML environments (AWS, GCP, Azure) and ML pipelines.

Responsibilities

• Architect end-to-end automated labelling pipelines using Snorkel/ Cleanlab or active learning loops.
• Build HITL systems with interfaces for high-uncertainty samples.
• Implement QA and denoising to fix mislabeled data.
• Integrate labelling tools with data lakes and ML training infra.
• Fine-tune teacher models to generate high-quality pseudo-labels.
• Maintain data prep infra for data quality and speed.
• Perform data visualization and in-depth feature analysis to drive insights.
• Collaborate with Data Scientists, Software Engineers, and Product teams.

Skills

Python

PyTorch

TensorFlow

NumPy

Pandas

Scikit-learn

Active Learning

Weak Supervision

Data Visualization

Jupyter

Data Pipelines

Cloud ML (SageMaker, Vertex AI, Azure)

SQL/NoSQL

DVC

Labelling Tools

Education

Bachelor’s or Master’s degree in Computer Science, Engineering, Mathematics, Data Science

Tools

Snorkel

Cleanlab

AWS SageMaker Ground Truth

Vertex AI

Azure ML

DVC

Tableau/Power BI

Jupyter

Job description

Dyson seeks a Lead Data Intelligence ML Engineer to design automated data-labelling pipelines, reducing manual annotation via Active Learning and Weak Supervision. You will bridge data collection and model-ready datasets, ensuring high-quality labels at scale.

You will collaborate with Dyson’s global engineering teams and external partners, shaping data strategies, and building robust data prep and MLOps infrastructure in a fast-moving environment.