Company Description
ByoSync OS Layer is an advanced biometric authentication platform that replaces traditional OTPs and passwords with secure, real-time facial verification. Tailored for India's dynamic digital ecosystem, ByoSync operates seamlessly on any smartphone, utilizing encrypted tokens to ensure that no biometric data is stored. The platform enables secure payments, identity access, and digital services through a single scan, delivering a fast, private, hardware-free, and security-first authentication experience.
Role Description
We are looking for a Machine Learning / Computer Vision Engineer with a strong mathematical foundation and experience building CNN-based vision systems.
This is a fixed-term role for 2 months, with the opportunity to convert to a full-time position based on performance and business requirements.
You will work on face representation, geometric processing, normalization, model training, evaluation, and on-device optimization, contributing directly to production-grade ML systems deployed on resource-constrained devices.
What You'll Work On
- Design face representation and processing pipelines using geometry, linear algebra, and statistical modeling
- Develop robust normalization and alignment techniques across pose, scale, and illumination variations
- Build and train CNN-based models for face representation and liveness detection
- Formulate matching, scoring, and confidence estimation algorithms
- Create evaluation frameworks including accuracy, error rates, latency, and device-level analysis
- Optimize models for on-device execution, including:
- Quantization, pruning, and distillation
- Memory, latency, and power constraints
- Collaborate with mobile engineers to ensure real-time, production-ready inference
Qualifications
- Strong background in Computer Vision, Machine Learning, or Applied Mathematics
- Deep understanding of CNN architectures and representation learning
- Excellent grasp of linear algebra, probability, and optimization
- Experience designing custom normalization, alignment, or feature extraction pipelines
- Proficiency in Python and PyTorch (or similar frameworks)
- Experience training and validating models on real-world data
Nice to Have
- Experience with face analysis systems (recognition, liveness, identity verification)
- Knowledge of model compression and deployment on constrained devices
- Exposure to robust or security-sensitive ML systems
- Experience in fintech, identity, or other regulated domains
