Senior AI/ML Engineer (Smart City / Computer Vision)

Project Overview

Join a high-impact Smart City project in Saudi Arabia (Riyadh). You will be optimizing and creating machine learning models to process video streams from surveillance cameras. Due to strict security restrictions, there is no remote access to the environment or data.
• Type: 6-month contract
• Format: On-site presence in Riyadh, Saudi Arabia for the majority of the time
• Relocation/Travel Support: Visa, airfare, accommodation, and a daily allowance are fully provided.

Requirements

• Experience: At least 5 years of relevant AI/ML or computer vision engineering experience.
• Domain Expertise: Proven track record of delivering production video analytics in smart-city, public-space, transport, or comparable environments.
• Core Tech Stack: Advanced hands-on Python skills, PyTorch and/or TensorFlow, and OpenCV.
• Infrastructure & Tools: Linux, Git, Docker, GPU-based training/inference, and model serving.
• Video Analytics: Practical understanding of video streams, camera/VMS integration, asynchronous/concurrent processing, and time/location alignment.
• Soft Skills: English proficiency (effective communication), ability to work independently, lead the ML workstream, and coordinate interfaces with UI/UX and platform teams.

Responsibilities

• Model Development: Train, fine-tune, and evaluate computer vision models for object detection, counting, multi-object tracking, and classification (people and vehicles).
• Advanced Analytics: Develop crowd analytics (density estimation, occupancy, heatmaps) and historical forecasting trend models (footfall).
• Security & Anomalies: Implement event and anomaly detection for video alerts (overcrowding, restricted-area intrusion).
• Pipeline Optimization: Optimize models and inference pipelines for real-time processing on customer-provided GPU/edge hardware (quantization, batching, acceleration).
• Integration: Package models into deployable inference services and integrate them with video management interfaces (RTSP) and data pipelines (Vantiq).
• Engineering Quality: Maintain production-quality Python code, write automated tests, handle input validation, and manage structured logging/error handling.
• Data & MLOps: Define annotation guidelines, validate labels, handle model versioning, manage containerized deployments, and monitor data drift.

Key Deliverables

• Trained and validated models with architecture details, weights, and training/inference scripts.
• Baseline and final evaluation reports using representative test data.
• Complete ML source code, automated tests, and reproducible experiment records.
• Deployable container images, build definitions, and troubleshooting runbooks.
• Technical support and knowledge transfer during SIT, UAT, and production deployment.