Data Scientist, KYC Anti Fraud

Job Overview

At our organization, Binance, we are inviting individuals to join our dynamic team as a Data Scientist. The focus of this role is on contributing to the Risk Data Science unit and actively participating in the development of in-house KYC anti-fraud solutions. The primary responsibilities encompass the creation and enhancement of AI/ML models tailored for identifying identity fraud, particularly concentrating on ID forgery detection, document recognition, and image-based analysis. In this fast-paced fintech atmosphere, collaboration with various teams such as Risk, Engineering, Product, and Operations will be essential to successfully deploying fraud detection solutions.

Key Responsibilities

• Create and manage in-house KYC and anti-fraud solutions, specializing in ID forgery detection, document recognition and verification, and image and video-based fraud analysis.
• Formulate, train, optimize, and assess computer vision or LLM models for fraud detection purposes incorporating aspects such as image quality assessment, tampering and forgery cues, and signals linked to AI-generated content or deepfakes.
• Examine extensive datasets with possible patterns indicative of suspicion to build features for model training and manual scrutiny.
• Collaborate with engineering counterparts to implement models into production pipelines and coordinate with Strategy and Operations teams to validate results and enhance detection capabilities.
• Oversee model performance post-production launch and continuously refine to achieve improved coverage whilst preserving low false-positive rates.

Requirements

• Possess a Bachelor’s Degree or higher in Computer Science, Data Science, AI, or a related field.
• Demonstrate proficiency in Python and commonly used data science / ML libraries.
• Exhibit a solid grasp of CV/ML/AI core concepts.
• Have experience with various computer vision techniques like image preprocessing, feature extraction, OCR, etc.
• Familiarity with model training, fine-tuning, and evaluation methodologies.
• Capability to work with large datasets and conduct exploratory data analysis effectively.
• Showcase strong problem-solving skills and the ability to operate autonomously in a dynamic setting.

Preferred Qualifications

• Prior involvement in KYC, fraud detection, or risk-related sectors, with hands-on experience in ID forgery detection, OCR and document interpretation, and image tampering or forgery analysis.
• In-depth familiarity with applying or refining LLMs for analysis, categorization, or automation assignments, and dealing with unlabeled or imbalanced datasets.
• Proficiency in image quality assessment (IQA), frequency-domain analysis, and AI-generated content or deepfake identification techniques.
• Understanding of AWS or cloud-based ML workflows along with platforms like SageMaker, and exposure to batch or online inference pipelines.
• Ability to present technical findings clearly to non-technical stakeholders.

Why Choose Binance

• Drive the future with the premier blockchain ecosystem globally.
• Collaborate with exceptional talent in a worldwide organization focused on users within a flat structured environment.
• Undertake innovative projects with liberty in a progressive surrounding.
• Flourish in an outcome-centric workplace with prospects for career advancement and ongoing learning.
• Avail competitive compensation packages and company benefits.
• Flexibility to work remotely as per business team requirements.

Binance vows to be an equal employment opportunity provider as we believe diversity in our workforce is fundamental to our success. Submission of an application indicates your acknowledgment and acceptance of our Candidate Privacy Notice. Artificial intelligence (AI) tools may be employed in parts of the hiring process to support application review, resume analysis, or response evaluation. While these tools aid our recruitment team, human judgment remains paramount in finalizing hiring decisions. Further information on data processing can be obtained upon contact.