Senior Data Scientist

Senior Data Scientist
Key Responsibilities
Design and develop state-of-the-art deep learning models for biometric recognition (facial, fingerprint, iris) and fraud detection using convolutional neural networks (CNNs), transformers, and hybrid architectures
Lead the design of end-to-end AI pipelines: data ingestion, pre-processing, model training, evaluation, and deployment in production environments
Conduct thorough error analysis, A/B testing, and performance evaluations to identify weaknesses and improve recognition accuracy and system efficiency
Architect scalable data pipelines for image and biometric signal cleaning, feature extraction, clustering, and deduplication at national scale
Collaborate with domain experts, biometric vendors, and legal/policy teams to ensure technological solutions are interpretable, robust, and policy compliant
Monitor system drift, adversarial threats, and fraud patterns in biometric usage across the Aadhaar ecosystem
Publish internal technical reports and participate in peer-reviewed research where appropriate
Mentor and guide junior data scientists and ML engineers in best practices
Education
B.Tech / B.E. from a Tier-1 Institution (IITs, IISc, IIITs, BITS, NITs, or similar institutions), preferably in Computer Science, IT, Electronics Engineering, Electrical Engineering, Mathematics, or a related field
Experience & Skills
Minimum 6+ years of hands-on experience in machine learning, deep learning, or computer vision roles
Strong expertise in deep learning frameworks such as PyTorch, TensorFlow, and Keras
In-depth knowledge of CNNs, Transformers, Siamese networks, contrastive learning, and embedding space analysis
Proficient in Python and experienced with large-scale data processing tools like Spark, Dask, or Hadoop
Prior experience working with biometric or image datasets (face, iris, fingerprint) is desirable
Strong understanding of data cleaning, clustering (e.g., DBSCAN, K-Means), anomaly detection, and unsupervised learning
Knowledge of privacy-preserving techniques such as federated learning, differential privacy, and homomorphic encryption is a plus
Proven record of taking models from research into production at scale, with MLOps exposure (Docker, Kubernetes, MLflow, model monitoring and drift detection)
Experience on large-scale government or public digital infrastructure programmes is a plus
What We Offer
A unique opportunity to work on the largest biometric identity system in the world
Access to UIDAI’s unparalleled scale of identity data and secure research environments
Competitive compensation and benefits tailored for top-tier AI talent in the public sector
A mission-driven organization with direct impact on national-scale governance, security, and inclusion

Work arrangement
No