Principal / Staff Machine Learning Engineer — Biometrics & Anti-Spoofing

Location: Remote / Lisbon / Flexible

Our client is a global technology company specializing in biometric identity verification, fraud prevention and AI-powered security solutions.

Their technology helps banks, fintechs, identity verification providers, telecommunications companies and public institutions securely verify identities and onboard customers at scale.

Operating internationally, the company develops advanced biometric solutions designed to detect and prevent identity fraud, including deepfakes, presentation attacks, injection attacks and other increasingly sophisticated threats.

Their solutions operate in highly regulated environments, where security, accuracy, performance and compliance are critical. The company combines advanced machine learning research with real-world production systems, with its technology undergoing independent testing and certification against recognized international standards.

As fraud techniques become increasingly sophisticated, our client is investing in its machine learning capabilities to stay ahead of emerging threats.

Our client is looking for a Principal / Staff Machine Learning Engineer to take a leading technical role in developing advanced biometric security and anti-spoofing solutions.

This is a highly senior, hands-on position combining deep learning research, architecture ownership and production engineering.

The successful candidate will be one of the most senior technical contributors within the ML team, taking ownership of core model architectures and helping define the research direction for key product areas, particularly Presentation Attack Detection (PAD) and Injection Attack Detection (IAD).

The role offers significant technical autonomy, with the opportunity to influence product strategy, mentor other engineers and work closely with leadership on the technology roadmap.

This is not a research-only position. The expectation is to develop models that perform against real-world fraud, can be deployed at scale and meet rigorous independent evaluation standards.

• Design, train and deploy deep learning models for face liveness detection, Presentation Attack Detection (PAD), Injection Attack Detection (IAD) and deepfake or synthetic media detection.

• Develop state-of-the-art approaches to counter evolving adversarial threats, including generative AI models, 3D masks, deepfake injection, virtual cameras, emulators and face-morphing attacks.

• Own the full machine learning lifecycle, from data strategy and model architecture to training infrastructure, evaluation and production deployment.

• Build and optimize ML solutions for web, mobile (Android/iOS SDKs) and server-side environments.

• Define robust evaluation methodologies and support the preparation of models for independent testing and certification, including iBeta, BixeLab, NIST FRVT and ISO/IEC 30107-3 standards.

• Optimize models for real-world production constraints, including low-latency inference, mobile and edge deployment, hybrid architectures and adversarial robustness.

• Stay at the forefront of research in face anti-spoofing, deepfake detection, adversarial machine learning and biometric template protection, translating relevant advances into production-ready solutions.

• Provide technical leadership across the ML team, mentoring engineers and reviewing architecture, model design and research decisions.

• Collaborate closely with backend, mobile and product teams to deliver secure, scalable solutions that meet enterprise security and compliance requirements.

• 8+ years of hands-on experience in machine learning and deep learning, including a strong track record of deploying models into production at scale.

• Deep expertise in computer vision, particularly in one or more of the following areas: face recognition and verification, liveness detection, anti-spoofing, deepfake detection or adversarial robustness.

• Proven experience designing and training CNN- and transformer-based computer vision architectures from scratch, beyond simply fine-tuning existing models.

• Strong understanding of adversarial threats targeting biometric systems, including presentation attacks, injection attacks and GAN- or diffusion-based deepfakes.

• Demonstrated ability to take ML models from research through to production, including optimization for mobile and edge inference using techniques and frameworks such as quantization, distillation, ONNX, TensorRT or CoreML.

• Solid foundations in probability, statistics, optimization and model evaluation, including ROC/DET curves, APCER/BPCER and FAR/FRR metrics.

• Experience working with large-scale data pipelines, dataset curation and the responsible handling of sensitive biometric information.

• A track record of technical leadership, whether through leading complex ML projects, defining technical direction or mentoring other engineers.

• Excellent written and verbal communication skills in English.

The ideal candidate combines a strong research mindset with practical engineering expertise. They can independently explore complex technical problems, make architecture decisions and deliver robust machine learning systems that operate reliably in production.

• Previous experience in biometrics, identity verification, KYC/IDV or fraud prevention.

• Publications, patents or significant research contributions in face anti-spoofing, deepfake detection or adversarial machine learning.

• Experience preparing ML systems for independent security or compliance certification, including iBeta, BixeLab, NIST or ISO 30107.

• Familiarity with on-device machine learning and deployment through iOS and Android SDKs.

• Experience working with generative models, including GANs and diffusion models, particularly from a red-team or adversarial testing perspective.

This is an opportunity to work on some of the most challenging problems in applied computer vision and adversarial machine learning, where research translates directly into real-world impact.

The successful candidate will contribute to technology used to prevent identity fraud and protect digital transactions across financial institutions, enterprises and public-sector organizations.

The position offers:

• High technical autonomy and ownership of critical machine learning architectures and research direction.

• The opportunity to develop and deploy models that address real-world security threats at scale.

• Direct collaboration with technical leadership and meaningful influence over the product and technology roadmap.

• Exposure to advanced research challenges in deepfake detection, biometric security and adversarial ML.

• An international, distributed engineering environment.

• Competitive compensation aligned with seniority and technical expertise.

The role offers a flexible working arrangement, with remote opportunities and potential office-based collaboration depending on the candidate's location and the team's needs.

Candidates based in Lisbon or open to an international working environment are welcome.