Description

You will develop and optimize machine learning systems for document processing and image understanding.

Responsibilities

• Maintain and optimize the existing document code prediction system.
• Collaborate with cross-functional teams to implement system enhancements based on business requirements.
• Monitor system performance and troubleshoot technical issues.
• Research machine learning advancements to improve prediction models.
• Identify new machine learning opportunities within the document processing domain.

Required Skills

• 4+ years of experience in machine learning systems, preferably in document processing.
• Bachelor's degree in Computer Science, Machine Learning, or a related field.
• Proficiency in Python.
• Hands-on experience with PyTorch or TensorFlow.
• Expertise in Image Transformer models, specifically Microsoft's DiT for document image understanding.
• Experience implementing self-supervised learning techniques for pre-training models on large-scale unlabeled text images.
• Proven ability to apply Transformer models to document image classification and layout analysis.
• Experience leveraging self-supervised pre-trained models as backbones for downstream tasks.
• Skill in designing experiments, analyzing results, and fine-tuning models.

Preferred Skills

• Familiarity with integrating Transformer models into OCR pipelines for text detection and extraction.
• Solid understanding of image processing using OpenCV for resizing and feature extraction.
• Experience with AWS or Azure and Docker containerization.

Key Skills

Education

Bachelor's degree