Machine Learning Engineer OCR & Vision Models

Role Summary:

Builds and fine-tunes OCR models, handwriting models, document segmentation, stamp detection, and image cleanup pipelines.

Key Responsibilities :

• Develop, fine-tune, and optimize OCR and document segmentation models.
• Implement preprocessing: deskew, denoise, watermark/stamp removal, cropping, binarization.
• Train CV models (YOLO/SAM) for stamp detection, handwriting region detection.
• Evaluate multiple OCR engines and build benchmarking frameworks.
• Build orchestration pipelines enabling ensemble predictions.
• Optimize inference for large-scale throughput (millions of docs).

Qualifications:

• 3 -8 years in CV/ML.
• Deep knowledge of PyTorch, TensorFlow, CNNs, Vision Transformers, OpenCV.
• Prior OCR/DocAI experience preferred.