What You Will Do (Core Responsibilities):

Design, train, and deploy machine learning models for image-based fraud detection, including:  

Deepfake detection  

Injection attack detection  

Digital manipulation analysis in biometric verification  

Work end-to-end across the ML lifecycle:  

Dataset curation (large-scale, noisy, adversarial datasets)  

Model development and training  

Evaluation and iteration using fraud-relevant metrics  

Production deployment and monitoring  

Build robust data pipelines, including:  

Data validation, cleaning, and labeling strategies  

Handling class imbalance, bias, and distribution shift  

Define and execute evaluation frameworks focused on real-world performance:  

Precision/recall trade-offs  

False positive vs. fraud detection balance  

Robustness to unseen attack types 

Contribute to production ML systems, ensuring:  

Scalability and reliability  

Monitoring and performance tracking  

Continuous improvement against evolving threats   

Collaborate closely with:  Fraud & AI research teams, Data collection and annotation teams, MLOps and platform engineering, Product teams  

Who You Are (Soft Skills):

A pragmatic problem-solver who understands the gap between research and production  

Comfortable working in adversarial, fast-evolving problem spaces  

Able to clearly communicate technical concepts and trade-offs  

Collaborative and adaptable, with a strong sense of ownership  

Motivated by building technology that has real-world impact  

What You Need (Required Knowledge, Technical Skills):

Required Qualifications:

Bachelor’s degree in Computer Science, Engineering, or related field  

2+ years of experience deploying machine learning models into production  

Strong background in computer vision (image-based ML)  

Solid programming skills in Python  

Hands-on experience with PyTorch and/or TensorFlow  

Experience working with real-world datasets and building data pipelines  

Tech Stacks:

Cloud: AWS  

Languages: Python  

ML Frameworks: PyTorch, TensorFlow, Scikit-learn  

Data Tools: Pandas, OpenCV  

Infrastructure: Docker, CI/CD, cloud-based ML pipelines  

What Would be Nice (Preferred Experience):

Advanced degree (PhD or equivalent experience in Machine Learning or Computer Vision)  

Experience in fraud detection or adversarial ML domains  

Experience with deepfake detection, image forensics, or manipulation detection  

Familiarity with generative AI models (training or analysis)  

Background in data science or data engineering  

What We Provide (Benefits):

Competitive package 

Full Remote contract 

Annual Leave 

Home Office Allowance 

Annual Bonus – up to 10% 

Health Insurance 

Learning & Development: We promote continuous learning and support role-aligned development opportunities, with access to a complimentary LinkedIn Learning licence.