JOB DESCRIPTION
Lead the architecture, development, and deployment of AI solutions specializing in Computer Vision, OCR, and Intelligent Document Processing (IDP).
Research, develop, and optimize Deep Learning models for image recognition, object detection, classification, segmentation, and document processing.
Design and optimize end-to-end OCR pipelines, including image preprocessing, text detection and recognition, document layout analysis, table extraction, and structured information extraction.
Develop multilingual OCR solutions, particularly for Vietnamese, English, and Japanese documents.
Perform model training, fine-tuning, evaluation, and optimization to improve accuracy, inference speed, and CPU/GPU efficiency.
Design and deploy scalable AI services supporting batch processing, real-time inference, and enterprise system integration.
Establish MLOps practices for model versioning, performance monitoring, and continuous improvement.
Provide technical leadership, mentoring, architecture reviews, and code reviews for AI Engineers; establish reusable AI components and engineering standards.
Collaborate with BA, PM, Backend, Data Engineering, and DevOps teams throughout the solution delivery lifecycle.
Participate in requirement analysis, solution consulting, technical feasibility assessment, effort estimation, PoC development, and customer technical discussions.
REQUIREMENTS
Bachelor's degree or higher in Computer Science, Information Technology, Artificial Intelligence, or a related field.
Minimum 5 years of experience in AI/ML development, including at least 3 years specializing in Computer Vision/OCR and 2 years in a Technical Lead or equivalent role.
Strong proficiency in Python, OpenCV, and at least one Deep Learning framework such as PyTorch or TensorFlow.
Solid understanding of CNNs, Vision Transformers (ViT), Object Detection, Image Segmentation, Transfer Learning, and Fine-tuning.
Hands-on experience with models and frameworks such as YOLO, ResNet, U-Net, PaddleOCR, Tesseract, or equivalent technologies.
Strong understanding of OCR pipelines, text detection, text recognition, document layout analysis, table extraction, and Key Information Extraction (KIE).
Experience handling challenging OCR scenarios involving low-quality images, scanned documents, complex tables, and multilingual documents.
Experience in dataset preparation, annotation, augmentation, model training, and evaluation using appropriate metrics such as Precision, Recall, F1-score, mAP, CER, and WER.
Proven experience deploying at least one Computer Vision/OCR solution into production.
Experience with Linux, Docker, Git, REST APIs, and CPU/GPU inference optimization.
Strong capabilities in solution architecture, complex technical problem-solving, engineering mentoring, and code quality management.
Good English communication and technical documentation skills.
Experience with Japanese OCR, invoice processing, contracts, and enterprise document extraction.
Experience with ONNX Runtime, TensorRT, OpenVINO, Triton Inference Server, or MLflow.
Knowledge of Vision-Language Models (VLM), Document AI, and hybrid OCR/LLM solutions.
Experience working with Japanese or Singaporean customers, including presales, effort estimation, and technical proposal preparation.
BENEFITS
Minimum 13 months salary per year - not including other bonuses such as KPI bonus for work efficiency, project bonus and revenue bonus. We do performance review twice a year. You will work in a professional, dynamic and friendly environment.
VTI offer annual health check-ups and fully pay social insurance, health insurance and unemployment insurance premium following the Labor law.
We offer one vacation/company trip and 4 teambuilding trips per year for every employees, along with various entertainment activities including: Swimming Clubs, Yoga, Zumba, Kendo and music order via internal Radio channel.
Two MVPs will be rewarded with a free trip to Japan, Taiwan, Singapore, or else.
We offer variety promotion opportunities and chance to raising income for people with capacity, enthusiasm, and long-term commitment.
We offer free Japanese class at the company.
We provide training opportunities to help our people to improve their skills. We support our members to learn and get Cloud, AWS, PMF, PMP certification.
Working hour from: 08:30am to 05:30pm. From Monday to Friday.
