Senior AI Engineer, Vision (Edge / Cloud)

An international technology group building AI-driven security products is expanding its Vietnam engineering team. They are seeking a hands-on Senior AI Engineer to own computer vision R&D from prototype to commercial release. This role emphasizes deep expertise in either edge deployment on constrained hardware or building scalable AI pipelines in the cloud, utilizing modern architectures like CNNs, Vision Transformers, and generative techniques for synthetic data.

What You Will Do

• Design, train, and fine-tune deep learning models for demanding visual recognition problems, including object attribute detection and text recognition, using modern architectures such as CNNs and Vision Transformers.

• Take a data-centric approach to model quality, using generative techniques (diffusion models, GANs) to create synthetic data where real samples are scarce or hard to capture.

• Write clean, efficient, and well-documented production code in Python and/or C++.

• Move models from research prototypes into robust, shipping products.

Specialization 1: Edge Deployment

• Profile trained models, find latency bottlenecks, and tune them to run within tight compute and power budgets.

• Apply compression techniques including INT8/FP16 quantization and pruning.

• Convert and deploy models with TensorRT, TFLite, or ONNX on embedded targets (ARM CPUs, NPUs, low-power hardware).

Specialization 2: Cloud AI and Data Mining

• Design algorithms and AI pipelines that extract insight from very large, unstructured video and image datasets.

• Lead large-scale model training and optimize deep learning architectures for throughput.

• Integrate vision-language models to enable multimodal search and retrieval.

What You Bring

• Bachelor’s or Master’s degree in Computer Science, AI, Computer Vision, or a related field.

• 5+ years of applied experience in machine learning, computer vision, or AI R&D.

• A proven record of taking models from research into commercial production.

• Deep working knowledge of PyTorch or TensorFlow.

• Strong Python and/or C++ skills in a Linux environment.

• Demonstrated depth in one or both specializations above.

Nice to Have

• Experience with multimodal foundation models (e.g., CLIP, LLaVA).

• Experience with active learning or agent-based automation of data workflow

Due to the high volume of applications we are experiencing, our team will only be in touch with you if your application is shortlisted.