AI Solution Architect - Industrial AI

Role : AI Solution Architect - Industrial AI

Location : Riyadh, KSA

Experience : 10+ Years

Must-Have:-

• Designed, developed, and deployed Machine Learning and Deep Learning models for industrial challenges involving large-scale numerical, categorical, and time series data, as well as images and videos.
• Extensive experience in Python programming for statistical analysis, custom algorithm development, and business insight generation using ML/DL models.
• Proficient in optimization techniques including Genetic Algorithms, Linear and Quadratic Programming, and others
• Skilled in comprehensive data workflows: collection, exploratory analytics, data cleansing, feature selection, and model validation.
• Experience in deploying AI/ML solutions using MLOPs on cloud platforms such as Azure and AWS.
• Solid knowledge of NLP, Large Language Models, Retrieval-Augmented Generation, and ML algorithms such as Decision Trees, Clustering, Support Vector Machines, Artificial Neural Networks, LSTM, CNN, and YOLO, with awareness of their practical advantages and limitations.
• Motivated by continuous learning and mastery of emerging technologies and methodologies in the fields of artificial intelligence and machine learning.

Good-to-Have.

• Experience in Oil & Gas, refinery, asset monitoring or other process industries.
• Exposure to GenAI, AI agents, RAG, knowledge graphs or hybrid physics-ML modelling.
• Knowledge of MLOps platforms, drift detection and responsible AI practices.

Responsibilities / Expectations from the Role

1 - Develop AI/ML models for process deviation detection, equipment performance, predictive insights, yield estimation and optimization support.

2 - Work with the Refinery Process SME to convert engineering logic and operational scenarios into model features, rules and evaluation criteria.

3 - Perform data exploration, feature engineering, model training, validation, tuning and explainability analysis.

4 - Deploy models into real-time or on-demand workflows and integrate outputs with alerts, dashboards, reports and recommendation services.

5 - Implement model monitoring, versioning, retraining, drift detection and performance reporting across the model lifecycle.

6 - Document assumptions, datasets, validation results and limitations, and support user acceptance testing and production stabilization.