About the Role
Join the Samsung Food AI team to develop and deploy deep learning-based computer vision solutions that enhance user experiences across Samsung Food products. As a Senior Deep Learning Engineer, you will work on cutting-edge computer vision applications, including image recognition, detection, segmentation, and generative modeling. Your contributions will directly power intelligent features on mobile and embedded platforms. While Computer Vision is the primary focus of this role, responsibilities may also include developing Natural Language Processing (NLP) models where relevant. Experience with NLP or multimodal machine learning (e.g., combining visual and textual inputs) is considered a strong plus.
Tech Stack
- Python
- PyTorch
- PyTorch Lightning
- OpenCV
- TorchVision
- Transformers
- Nvidia Triton
- MLFlow
- DVC
- KubeFlow
- Kubernetes
- GCP
- ClickHouse
- MongoDB
- MySQL
Responsibilities
- Improve and optimize CV models for image classification, object detection, segmentation, and image generation.
- Conduct R&D using our multimodal data (recipe images, videos, user-generated content).
- Stay up to date with the latest developments in CV and DL research and apply them to production use cases.
- Develop scalable, efficient inference pipelines for real-world applications across mobile and server environments.
- Collaborate closely with ML engineers and data engineers to improve model serving infrastructure and ML pipelines.
- Support continuous integration and monitoring of ML models in production environments.
- Collaborate on NLP-related tasks where applicable, particularly in multimodal or content-driven applications.
Requirements
- Minimum 4 years of experience in a machine learning engineering role.
- Demonstrated experience in building and deploying computer vision models in production.
- Strong theoretical and practical knowledge of CV techniques and architectures (e.g., CNNs, Vision Transformers, YOLO, UNet, Diffusion Models).
- Strong hands-on experience with Python and PyTorch (our primary DL framework).
- Experience with image recognition, segmentation, object detection, or visual classification tasks.
- Familiarity with model optimization techniques (e.g., quantization, pruning, distillation).
- Understanding of trade-offs in performance and latency for real-world applications, including deployment to mobile and server environments.
- Solid software engineering practices, including clean, maintainable, and production-grade code.
- Working proficiency in English.
Nice to Have
- Experience with on-device CV model deployment (e.g., TensorRT, TFLite, CoreML, or other edge inference frameworks).
- Knowledge of embedded ML/AI on ARM or mobile platforms (e.g., Android NNAPI, iOS).
- Exposure to multimodal learning, particularly combining visual and textual inputs.
- Experience with NLP models and tasks (e.g., classification, entity extraction, text embeddings).
Location
Fully Remote: Work from anywhere with flexibility, though overlap with GMT time-zones is ideal.