Applied Scientist I (AI)

TL;DR

Applied Scientist I (AI): Designing and training machine learning solutions for digital identity security with an accent on deepfake detection, bias mitigation, and document understanding. Focus on implementing fairness strategies, optimizing model training workflows, and improving large-scale computer vision products.

Location: Must be based in the UK (Hybrid: 3 days per week in London office).

Company

A global industry leader in identity-centric security solutions, providing AI-powered verification technologies to over 150 countries.

What you will do

• Push the frontier of research in areas such as deepfake detection, bias mitigation, and document understanding.
• Implement and benchmark large-scale vision-language models using synthetic and real-world data.
• Profile, debug, and improve model training speed on distributed GPU systems.
• Collaborate with product and engineering teams to deploy world-class identity-focused models.
• Publish research findings in international conferences and scientific journals.

Requirements

• Strong academic or practical experience in machine learning and computer vision.
• Deep understanding of machine learning theory.
• Strong coding skills in Python and PyTorch.
• Proven track record of delivering high-performance ML-driven products.
• Must be based in the UK for a hybrid working arrangement.

Nice to have

• Technical experience in face matching, anomaly detection, or on-device ML.
• Experience optimizing distributed training code.
• Published papers at top-level machine learning conferences.

Culture & Benefits

• Focus on diversity, inclusion, and professional growth through learning-forward initiatives.
• Collaborative environment with support from dedicated ML Ops teams.
• Access to state-of-the-art tooling including AWS, Encord, Ray, and Weights & Biases.
• Flexible work options to support work-life balance.