Marcos Santos
Vice President Of Research And Development at ZeroEyes
United States
Summary
🤩
Rockstar🎓
Top SchoolMarcos Santos is a Senior Computer Vision Engineer with seven years of experience building production-grade vision and AI systems for multi-camera, edge and cloud deployments. Trained in mechatronics and control automation, he blends electronics and low-level programming expertise in C/C++ and Python to optimize models and pipelines for NVIDIA (CUDA, TensorRT, DeepStream, Jetson) and AMD (ROCm, HIP) ecosystems. He’s delivered solutions across object detection, pose estimation, segmentation, OCR and multimodal reasoning using YOLO, DETR, VLMs, LLMs, PyTorch and OpenVINO, and actively contributes performance-focused open-source work such as DeepStream-Yolo with custom CUDA kernel optimizations for Jetson devices. Known for turning research prototypes into robust automation, he pairs system-level pragmatism with a knack for squeezing edge hardware for real-time inference.
7 years of coding experience
6 years of employment as a software developer
Graduação Mechatronics, Robotics, and Automation Engineering, Graduação Mechatronics, Robotics, and Automation Engineering at Universidade Salvador
English, Portuguese
Stackoverflow
11reputation
318reached
0answers
1question
Github Skills (13)
tensorrt10
cuda10
kernel10
object-detection10
nvidia-jetson9
cprogramming-language9
c-language9
computer-vision9
darknet8
model-optimization8
objective-c6
e-mail6
ios6
Programming languages (5)
Github contributions (5)
marcoslucianops/DeepStream-Yolo
Dec 2020 - Jan 2023
NVIDIA DeepStream SDK 8.0 / 7.1 / 7.0 / 6.4 / 6.3 / 6.2 / 6.1.1 / 6.1 / 6.0.1 / 6.0 / 5.1 implementation for YOLO models
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:134 commits, 16 PRs, 530 pushes in 2 years 1 month
Contributions summary:Marcos primarily contributed to the implementation and modification of custom CUDA kernels for YOLO object detection models, focusing on optimization and bug fixes related to NVIDIA Jetson Nano. They worked on CUDA code, including modifications to the `yoloForward.cu` file and integration within the DeepStream SDK environment. The commits demonstrate a focus on integrating and fine-tuning YOLO models within the DeepStream framework, specifically for edge devices.
deepstreamnvidianvidia-deepstream-sdkobject-detectionyolo
marcoslucianops/DeepStream-Yolo-Pose
Sep 2023 - Oct 2025
NVIDIA DeepStream SDK 8.0 / 7.1 / 7.0 / 6.4 / 6.3 / 6.2 / 6.1.1 / 6.1 / 6.0.1 / 6.0 application for YOLO-Pose models
Contributions:1 PR, 17 pushes, 1 branch in 2 years 1 month
deepstreamnvidianvidia-deepstream-sdkobject-detectionpose-estimation
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.