AI Engineer — Vision AI & Edge Intelligence

Syanxg Technologies Pte Ltd; Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

Syanxg Technologies Pte Ltd is seeking a Vision AI Engineer to design and deliver production-grade Vision AI systems for edge and cloud use. You will work on object detection, segmentation, tracking, and video analytics, while optimizing inference pipelines with TensorRT/ONNX. Collaboration across teams will drive commercial Vision AI solutions.

You will fine-tune vision-language models for domain-specific tasks and help deploy end-to-end AI solutions from model to production rollout.

Qualifications

• Strong background in computer vision, deep learning, and AI application development.
• Proficiency in PyTorch, TensorFlow, or equivalent frameworks.
• Hands-on experience with object detection, segmentation, tracking, or video analytics.
• Knowledge of MLLMs, VLMs, and multimodal model adaptation.
• Experience deploying AI models with ONNX or TensorRT.
• Experience with NVIDIA Metropolis, DeepStream, or similar stacks.
• Proven production deployment of Vision AI solutions in commercial environments.
• Strong Python skills and familiarity with C++ for AI apps.

Responsibilities

• Develop computer vision models for object detection, segmentation, tracking, and video analytics.
• Implement and optimize Vision AI algorithms for edge and cloud deployments.
• Fine-tune vision-language models for domain-specific tasks.
• Build Vision Agent capabilities for live or recorded video streams.
• Create low-latency inference pipelines using TensorRT/ONNX Runtime or equivalents.
• Support end-to-end AI deployment from integration to production rollout.
• Collaborate with teams to deliver production-grade Vision AI solutions.

Skills

Vision AI development

Deep learning

PyTorch/TensorFlow

Object detection

Video analytics

MLLM/VLM adaptation

Production deployment

Python programming

C++ for AI

Education

Bachelor's degree or higher in CS/AI/EE

Tools

Docker

Kubernetes

CUDA

ONNX/TensorRT

Job description

Key Responsibilities

Vision AI Development & Edge Real-Time Inference

• Develop computer vision models for object detection, segmentation, tracking, video analytics, and scene understanding.

• Implement and optimize Vision AI algorithms for commercial applications across diverse edge deployment scenarios.

• Fine-tune and adapt vision-language models (VLMs) and multimodal large language models (MLLMs) for domain-specific image and video understanding use cases.

• Develop Vision Agent capabilities that combine perception models, multimodal reasoning, natural-language interaction, and tool-based workflows for live or recorded video streams.

• Build and optimize low-latency AI inference pipelines using TensorRT, ONNX Runtime, or similar production inference frameworks.

• Support end-to-end AI deployment, including model integration, validation, performance tuning, and production rollout.

Required Skills & Experience

• Strong background in computer vision, deep learning, and AI application development.

• Proficiency in PyTorch, TensorFlow, or equivalent deep learning frameworks.

• Hands-on experience with object detection, segmentation, tracking, video analytics, or related Vision AI applications.

• Knowledge of MLLMs, VLMs, and foundation model adaptation for image, video, and text-based multimodal understanding.

• Practical experience deploying AI models using ONNX, TensorRT, or similar inference optimization frameworks.

• Hands-on experience with NVIDIA Metropolis and related vision AI deployment stacks such as DeepStream, TensorRT, or GPU-accelerated video analytics pipelines.

• Proven experience developing and deploying commercial Vision AI solutions in production environments.

• Strong programming skills in Python and familiarity with C++ for AI application development.

Preferred Skills

• Experience with multimodal learning, vision-language models (VLMs), multimodal large language models (MLLMs), or foundation model adaptation.

• Familiarity with Vision Agent concepts such as video question answering, video summarization, visual grounding, multimodal RAG, agentic tool use, and real-time interaction with video streams.

• Experience with NVIDIA AI software stack, including CUDA, TensorRT, DeepStream, or related GPU acceleration technologies.

• Familiarity with NVIDIA Isaac, NVIDIA Cosmos, and NVIDIA Omniverse for robotics, Physical AI, simulation, synthetic data, or digital-twin workflows is desired.

• Experience with Docker, Kubernetes, or containerized AI deployment.

• Background in real-time video processing, telecom systems, robotics, smart surveillance, autonomous systems, or Physical AI applications.

Education & Qualifications

• Bachelor's degree or higher in Computer Science, Artificial Intelligence, Electrical Engineering, or a related technical field.

• Strong foundation in machine learning, deep learning, computer vision, or applied mathematics.

Experience Requirements

• Minimum 3 years of relevant industry experience in AI, Machine Learning, or Computer Vision.

• Demonstrated experience developing and deploying production-grade Vision AI systems for commercial applications.

• Experience delivering AI solutions from model development through deployment and production support is an advantage.