AI/ML Engineer

AI / ML EngineerDubai About the Company: Our client is an AI and robotics startup building autonomous systems for heavy duty environments. Based in the UAE, they design, build and deploy intelligent autonomy end-to-end, helping enterprise clients bring automation into complex, high risk operations.The RoleYou will own how the vehicles see and understand the world. That means building the perception layer that turns raw sensor data into a reliable picture of the environment in real conditions: dust, glare, heat, uneven terrain and places no one has mapped. Your work feeds directly into navigation, obstacle avoidance and fleet management. This is a hands-on role with high ownership, and you will be expected to get code running on real vehicles early.What You Will DoDesign, build and deploy the perception stack across camera, LiDAR and radar inputsDevelop object detection, tracking and segmentation that holds up in harsh outdoor and industrial settingsBuild localisation and mapping capability, including operation in GPS-denied and unmapped areasFuse multi-sensor data into a unified, real-time environmental model the planning stack can trustOptimise models for on-vehicle inference on edge hardware within strict latency and compute budgetsBuild data pipelines for collection, labelling, validation and retraining from fleet dataFeed perception insights into fleet-level analytics for mission planning and performance monitoringWork closely with autonomy, controls and hardware engineers to take features from test track to deploymentWhat You Bring3+ years building perception systems that have run on real robots or vehicles, not only curated datasetsStrong computer vision skills: detection, tracking and segmentation on live sensor feedsHands-on experience with SLAM, point cloud processing and multi-sensor calibrationExperience with sensor fusion (e.g. Kalman filtering, factor graphs or learned fusion approaches)Strong Python and C++, Experience deploying models to edge or embedded GPU hardwareComfort with ROS/ROS 2 and field testingOwnership mindsetNice to HaveModel compression and optimisation (quantisation, pruning, TensorRT or similar)Experience with fleet data infrastructure and MLOps for continuous model improvementExposure to simulation environments for perception testingWhy ApplyBuild technology that runs on real vehicles in real missionsEarly-stage ownership in a fast-growing startup