Serge Alhalbi

About

Robotics software engineer with deep ROS2 experience who owned the perception stack for an automated C. elegans manipulation system—building YOLO-based worm segmentation plus OCR label reading and integrating it into a MoveIt2 pipeline with real-time latency constraints. Also deploying ROS2 on an AgileX Tracer with ZED depth camera for vision-based person following and working on SLAM/sensor fusion, with additional production-style ML deployment experience (Dockerized FastAPI + PyTorch on AWS EC2 with CI/CD).

Experience

Education

Key Strengths

• Owned and delivered an end-to-end robotics perception pipeline integrated into ROS2/MoveIt2 closed-loop manipulation
• Improved real-world robustness of YOLO segmentation via dataset curation, augmentation, and labeling iteration under challenging lab conditions
• Optimized perception latency for near-real-time robotics using CUDA acceleration, inference prewarming, and lightweight ROS interfaces
• Increased OCR reliability by standardizing label design and adding dual Luhn checksum validation to prevent bad reads propagating downstream
• Debugged and stabilized manipulation planning/execution by restructuring MoveIt Task Constructor logic and tuning planning constraints
• Strong ROS2 systems integration across perception, navigation, and manipulation (nodes/packages, tf2 frames, QoS, DDS, namespaces) on real hardware

Reference Highlights

• Proficient programmer
• Highly capable of learning quickly to fill technical gaps
• Strong debugging skills (code and data)
• Able communicator
• Works well in applied engineering contexts
• Self-directed; willing to figure things out independently
• Requires minimal supervision
• Strong potential for future growth
• Would be rehired
• Good at Python
• Good at ROS/ROS 2
• Excellent debugging skills
• Solved many issues on the first prototype system
• First to get the ROS 2 system running on the lab computer despite no prior team experience
• Outstanding at working with hardware teams
• Strong hardware design capability
• Designed the entire system from scratch (grounding work)
• Designed an end effector for proper functioning from scratch
• Well trained in both software and hardware skills