Research Engineer / Deep Learning Engineer (Computer Vision) – CDI

As a senior member of the R&D team, you will design, train, optimize, and deploy deep learning models for ShareID’s core products: Document Verification, Face Authentication, Liveness, and Fraud Detection. You will work on computer vision problems involving images, videos, and temporal sequences in highly adversarial (fraud-prone) environments.

Responsibilities

• Design and implement advanced computer vision models, with a focus on:
  • identity document analysis
  • forgery / tampering detection / Document spoofing
  • video-based temporal modeling and tracking
• Experiment with state-of-the-art architectures:
  • transformer-based models (ViT, DETR-like, SAM, etc.)
  • diffusion / generative models for augmentation or anomaly detection
  • latency-optimized networks (quantization, pruning, distillation)
• Own end-to-end research cycles: literature review, prototyping and experimentation, evaluation on large-scale datasets, productization with engineering teams.
• Collaborate cross-functionally with Product, Risk, Fraud, and Engineering to bring research ideas into production.
• Contribute to ShareID’s scientific culture: present papers, lead knowledge-sharing sessions, guide junior ML engineers and interns, optionally participate in benchmarks or publications.

Requirements

Minimum 4 years of experience in deep learning applied to computer vision, with a portion of that experience in a production environment (startup, scale-up, industrial lab, etc.).

• Excellent command of:
  • Python
  • PyTorch (or equivalent)
  • Large-scale model training (voluminous datasets, data augmentation, rigorous validation)
• Concrete experience in at least one of these areas:
  • Real-time vision (tracking, video detection, high-performance pipeline)
  • Document understanding (document scanning, OCR, document augmentation, QA)
• Solid foundation in:
  • Statistics, optimization, supervised / self-supervised learning
  • Good reading comprehension of literature (ICCV, CVPR, NeurIPS, etc.)
• Experience working in a product environment:
  • Latency, robustness, hardware resource, security, and privacy constraints