Head of Data Science - Document Verification & Biometrics

Socure Socure · Vertical AI · United States · Remote · Data Science & AI

Lead the data science strategy for Document Verification & Biometrics, evolving from computer vision to foundation model-driven, agentic architectures. Focus on building intelligent agents for verification workflows, advancing fraud detection (deepfakes, presentation attacks), and leveraging vector databases. This role requires expertise in computer vision, multimodal AI, foundation models, agentic systems, and customer-facing communication.

What you'd actually do

  1. Own and define the data science vision for Document Verification, Biometrics, and Reusable ID, delivering step-function improvements in accuracy, speed, and user experience.
  2. Lead and scale a high-performing team of data scientists and applied researchers, setting a high standard for execution, innovation, and accountability.
  3. Architect and deploy state-of-the-art computer vision and multimodal systems, including vision-language models (VLMs), for document understanding, face matching, liveness detection, and identity verification.
  4. Drive the development of domain-specific foundation models tailored to identity, documents, and biometrics, leveraging large-scale proprietary datasets.
  5. Lead the transition to agentic systems, building intelligent agents that can reason over document and biometric signals, automate verification workflows, and adapt dynamically to new fraud patterns.

Skills

Required

  • Advanced degree (MS/PhD preferred) in Computer Science, Electrical Engineering, Machine Learning, or a related field.
  • 10+ years of experience in data science, machine learning, or applied AI, with a strong track record of building and deploying production systems.
  • Deep expertise in computer vision and multimodal AI, including experience with vision-language models (VLMs).
  • Proven experience building domain-specific foundation models or large-scale representation learning systems.
  • Strong understanding of biometric systems, including face recognition, liveness detection, and anti-spoofing techniques.
  • Experience detecting and mitigating deepfakes, presentation attacks, and counterfeit documents.
  • Strong understanding of agentic system design, including agent skills, orchestration/harness frameworks, and real-world deployment of autonomous or human-in-the-loop agents.
  • Experience with vector databases and embedding-based retrieval systems.
  • Strong awareness of emerging attack vectors, including prompt injection, adversarial inputs, and model exploitation techniques.
  • Proven ability to lead and scale high-performing teams while delivering under ambiguity and tight timelines.
  • Proficiency in Python and modern ML frameworks (e.g., PyTorch).
  • Proven ability to engage with customers and external stakeholders, clearly explaining complex AI systems, trade-offs, and outcomes to both technical and non-technical audiences.

Nice to have

  • Experience representing organizations in customer-facing discussions, executive briefings, or public speaking engagements is strongly preferred.

What the JD emphasized

  • foundation model-driven, agentic architectures
  • domain-specific foundation models
  • agentic systems
  • multimodal systems
  • vector databases and embedding systems
  • deepfake detection
  • presentation attack detection
  • counterfeit documents
  • prompt injection
  • adversarial attacks

Other signals

  • foundation model-driven, agentic architectures
  • domain-specific foundation models
  • agentic systems
  • multimodal systems
  • vector databases and embedding systems