Full-time | On-site with flexibility
At Global ID, we develop finger-vein biometric identification systems, combining our own near-infrared scanners, matching algorithms and identity platform, deployed on-premises in real-world environments, including public-health programmes.
We are looking for a Senior Biometrics Engineer to take ownership of our entire biometric pipeline, from image quality assessment to large-scale identification.
You will improve our existing algorithms, lead the development of their successors, build robust evaluation methods and ensure reliable integration into our production platform.
This is a hands‑on senior role with genuine technical ownership, working closely with our scanner engineer, platform team, field operations in Kenya and academic research partners.
- Own the biometric algorithm roadmap: improve our current C++/OpenCV matcher and lead the development of future versions.
- Lead biometric evaluation: design rigorous protocols covering FMR/FNMR, DET curves, FPIR/FNIR, confidence bounds and cross-device or cross-session analysis.
- Build and manage datasets: define collection protocols, quality controls, labelling, consent requirements and data-protection safeguards.
- Improve end‑to‑end accuracy: work on image quality, capture feedback, image enhancement, feature extraction, learned matchers, two‑finger score fusion and threshold calibration.
- Design safe algorithm transitions: implement versioned templates and parallel matcher versions, enabling enrolled users to migrate progressively without disrupting existing deployments.
- Integrate algorithms into the product: collaborate on matcher services, APIs, Java/Python services, containers and release packaging.
- Drive performance and scalability: optimise processing times, memory usage and CPU performance, and develop search strategies for 1:N identification at scale.
- Lead research collaborations: define data formats, evaluate research deliverables and integrate promising results into the product.
- Document and share knowledge: produce evaluation reports, technical decision records and onboarding material.
- Applied biometrics or computer vision experience: several years working on systems used in real‑world applications. Around five years or more is a useful benchmark, but practical achievements matter most.
- Strong biometric evaluation skills: understanding of operating points, verification versus identification, false‑match risks and statistical confidence.
- Python and machine learning: strong Python skills for data analysis and model training, using PyTorch or an equivalent framework.
- C++ and image processing: ability to read, profile and modify image‑processing code, particularly with OpenCV.
- Real‑world dataset expertise: experience with imperfect datasets, augmentation, cross‑validation, leakage prevention and device/session variability.
- Production software engineering: experience with Java or Python back ends, REST APIs, Docker, testing and performance optimisation.
- Reproducible engineering practices: version control, containers, measurable improvements and reproducible pipelines.
- Autonomy and communication: ability to make technical decisions, document them clearly and flag issues early.
- Languages: professional English. French is an advantage, but not required.
- Work authorisation: right to work in Switzerland or eligibility to obtain it.
- Experience with vascular biometrics, particularly finger‑vein or palm‑vein recognition, or other near‑infrared imaging technologies.
- Experience with large‑scale identification systems, embeddings, approximate search, indexing, distributed matching or GPU inference.
- Knowledge of presentation attack detection, biometric template protection (ISO/IEC 24745), biometric standards or certification.
- Experience deploying technology in low‑resource environments, including public‑health programmes in emerging markets.
- Experience translating academic research into production‑ready solutions.
- Familiarity with modern AI‑assisted development tools.
- Real‑world impact: your measurements and decisions can influence field deployments within months.
- Full technical ownership: take responsibility for the biometric development line in a small, collaborative team.
- Technology in action: contribute to systems already in use, including public‑health pilots.
- Hardware and software under one roof: work directly with our scanner engineers to improve image capture at its source.
- Research‑driven environment: join us at EPFL Innovation Park, close to the academic and scientific community.
We would love to hear from you.
#J-18808-Ljbffr
