Edge ML Engineer – Adaptive Security for OT Systems

DISCOVERED; Dubai; On-site; AED 300,000 - 550,000; Full time

Job summary

Forge is seeking an hands-on Machine Learning Engineer to own the classical ML powering the detection and triage layer for OT environments. You will build models that score alerts, classify assets, and separate real threats from noise, running locally on air‑gapped appliances.

You will design adaptive ML systems that learn on the edge, evolving over time based on traffic patterns and analyst feedback, and you’ll stand up a self‑hosted MLOps stack for experiment tracking and model management.

Qualifications

• 4+ years of hands-on experience building and shipping ML systems in production.
• Strong Python fluency with XGBoost, LightGBM, or CatBoost.
• Solid grounding in scikit-learn, tree ensembles, clustering, anomaly detection (Isolation Forests, autoencoders).
• Experience with online/incremental learning, concept drift, and active learning loops.
• Proficiency with ML pipelines and MLOps tools like Airflow, Kubeflow, DVC, MLflow.

Responsibilities

• Alert triage and scoring: build and tune gradient boosting models to rank security alerts and surface genuine incidents.
• Adaptive and online learning: design models that adapt over time with incremental learning and human-in-the-loop feedback.
• Asset and device classification: profile OT devices from raw network data using tree ensembles and clustering.
• Custom anomaly detection: implement Isolation Forests and autoencoders with OpenSearch for detecting unseen threats.
• Risk scoring: refine features to reflect real operational exposure.
• Production pipelines: create reproducible, versioned feature engineering and training pipelines.

Skills

Python

XGBoost

LightGBM

CatBoost

scikit-learn

Incremental learning

Tools

MLflow

OpenSearch

Airflow

Kubeflow

DVC

Dagster

Prefect

Job description

Forge is seeking an hands-on Machine Learning Engineer to own the classical ML powering the detection and triage layer for OT environments. You will build models that score alerts, classify assets, and separate real threats from noise, running locally on air‑gapped appliances.

You will design adaptive ML systems that learn on the edge, evolving over time based on traffic patterns and analyst feedback, and you’ll stand up a self‑hosted MLOps stack for experiment tracking and model management.