Software Engineer, Perception (Robotics)

Location: Shanghai, China; onsite at a hirify.global office

Company

hirify.global is Southeast Asia's leading superapp, using technology and AI across delivery, mobility, and financial services.

What you will do

• Develop multi-object tracking with data association, Bayesian state estimation, track lifecycle management, and robust identity tracking through occlusions and detector gaps.
• Build open-set world understanding, including class-agnostic obstacle detection, occupancy representations, clustering, and generic-object detection.
• Develop motion prediction for pedestrians, cyclists, vehicles, and other agents using multi-modal forecasting, interaction-aware prediction, occupancy flow, and uncertainty estimation.
• Evaluate temporal perception-prediction systems, VLM/VLA and embodied foundation models, including grounding, long-tail reasoning, temporal consistency, latency, and robustness.
• Integrate camera, LiDAR, radar, and IMU data while maintaining geometric, temporal, identity, prediction, and semantic consistency.
• Define quality gates, investigate fleet-log regressions, and move systems from research through simulation, NVIDIA Orin/TensorRT deployment, and robot testing in partnership with engineering teams.

Requirements

• At least 3 years of hands-on experience shipping or operating perception and/or prediction systems for robotics, autonomous vehicles, or ADAS, including strong autonomous-driving experience and demonstrated real-world improvements.
• Rich experience with classical/geometric and learning-based approaches, including judgement on modular, end-to-end, and hybrid designs and interpretable safety fallbacks.
• Proven depth in multi-object tracking, including data association, Bayesian filtering, track lifecycle, multi-sensor/temporal association, ID stability, and low-latency evaluation.
• Hands-on experience with generic/open-set perception, such as class-agnostic obstacle detection, occupancy representations, LiDAR clustering, anomaly or unknown-object handling, or fusion with learned detections.
• Hands-on depth in at least one modern temporal area: motion forecasting, occupancy flow, streaming/temporal BEV, learned world models, or joint detection-tracking-prediction.
• Understanding of multi-modal uncertainty and the interface between perception, prediction, and planning.
• Fundamentals in geometry and multi-sensor systems, including practical calibration and diagnosis of alignment or time-synchronisation issues.
• Production-quality proficiency in C++ and Python, including training/evaluation tools and real-time embedded performance reasoning.
• Experience with reproducible experiments, clear metrics, log-driven root-cause analysis, simulation/replay, and validation-oriented execution.
• Self-motivated and dependable, with enthusiasm for autonomous driving, robotics, and embodied AI; demonstrated proficiency in using AI tools while maintaining quality, security, and reproducibility.

Nice to have

• End-to-end perception and prediction, including joint detection-tracking-forecasting, occupancy/occupancy-flow prediction, streaming BEV, trajectory queries, or world models.
• Motion prediction for autonomous systems, including multi-modal trajectories, interaction-aware forecasting, uncertainty calibration, planning interfaces, closed-loop simulation, or behaviour evaluation.
• VLM/VLA and embodied foundation-model experience, including vision-language grounding, open-vocabulary perception, task conditioning, semantic scene reasoning, long-tail mining, or action-relevant representations.
• Foundation-model adaptation and deployment using fine-tuning, LoRA, distillation, quantisation, teacher-student pipelines, edge inference, and systematic hallucination/grounding evaluation.
• 3D/BEV detection, sensor fusion, temporal models, 4D data, and multi-camera/LiDAR/radar integration.
• Robotics and perception platforms such as OpenCV, PCL, Ceres, ROS 2, NVIDIA Orin, TensorRT, or DeepStream, plus visualisation and debugging practices.
• Model-lifecycle experience covering dataset curation, auto-labeling, temporal annotation, active learning, regression detection, inference optimisation, and fleet-data feedback loops.

Culture & Benefits

• Work on urban embodied AI and autonomous delivery robots with hardware, control, machine-learning, and real-world fleet operations handled across the product lifecycle.
• Collaborate with Detection, Prediction, Planning, Data & Infra, and Integration teams.
• Work in a pragmatic research environment where measurable closed-loop value, reliable grounding, real-time performance, and safety determine production adoption.
• Contribute to robotics deployments across complex urban environments and large-scale last-mile logistics systems.