Risk Operations AI Specialist

About the Role

This position is part of a risk operations team undergoing an AI-led shift from manual case handling to a largely autonomous decision workflow powered by ROCA, the Risk Operations AI Agent. The role serves as a human review point for lower-confidence cases, where your assessments and annotations help refine the model and gradually widen automation coverage. It is a specialized operator role rather than a conventional fraud analyst position.

As the reviewer, you will interpret AI-generated signals, confirm or correct decisions, and provide the structured feedback needed to improve future model performance. The focus is on using technology to scale judgment, not on replacing expertise.

Core Review Duties

• Examine cases escalated by ROCA when the confidence score is below the auto-decision threshold, using the AI-supported review interface without switching between multiple systems.
• Decide daily case queues by approving, blocking, or escalating matters across KYC Audit, ATO investigations, Reset 2FA, POA Review, and withdrawal anomaly cases.
• Refer high-priority key user matters (Priority 75 and above) and elevated-risk patterns to a Senior Agent or Team Lead, with a clear written explanation for the escalation.
• Meet the manual resolution SLA of under 4 hours while maintaining decision accuracy of at least 98%.

AI Feedback and Pipeline Improvement

• Tag each reviewed case as true positive, true negative, false positive, or false negative, and add structured root-cause notes that feed the ROCA learning pipeline.
• Spot repeated false-positive and false-negative trends and send them through the appropriate feedback route: feature-weight bias for retraining or fine-tuning, threshold miscalibration for Decision Hub rule updates with A/B validation, or SOP/logic gaps for version-controlled SOP updates.
• Help improve precision above 95% and recall above 99% by identifying recurring signal failures and sharing them with Data and Model teams.
• Take part in model calibration reviews to inspect misclassified batches, confirm ground truth, and challenge decision thresholds using data.

SOP and Knowledge Management

• Keep Confluence SOP pages current for the assigned case types, aiming for at least 80% SOP coverage, since this directly affects ROCA first-pass accuracy.
• Identify gaps in SOP coverage whenever ROCA’s reasoning exposes unsupported scenarios, and prepare proposed SOP additions for Team Lead approval.
• Train and align BPO agents on AI-assisted review processes, and monitor their quality against automated QA standards.

Requirements

• At least 2 years of experience in fraud review, risk operations, or KYC/KYB case management.
• Practical experience investigating ATO, or account takeover, cases.
• Comfort operating in AI-augmented workflows, including reading model outputs, interpreting confidence scores, and making decisions from AI signals rather than raw data alone.
• Strong, precise written communication skills, with the ability to capture a root-cause annotation in fewer than 3 sentences in a form usable by an ML engineer.
• Bilingual English and Mandarin capability to coordinate with overseas partners and stakeholders.
• An operator-focused mindset toward AI tools, using machine assistance to make faster and better decisions.

Nice to Have

• Experience in crypto, Web3, or VASP risk environments, including awareness of blockchain activity patterns such as Sybil attacks, wash trading, and mule-account indicators.
• Exposure to model-evaluation concepts such as precision, recall, F1, and confusion matrices.
• Familiarity with prompt engineering or LLM evaluation workflows, especially reviewing or testing AI-generated decisions at scale.
• CAMS, CFE, or an equivalent AML/fraud certification.
• Previous experience writing SOPs, supporting QA calibration, or overseeing BPO teams in a risk operations function.

Why Binance

• Join a global blockchain ecosystem shaping the future of digital assets.
• Work with high-caliber colleagues in a user-focused organization with a flat structure.
• Take on fast-moving, high-autonomy projects in an innovative environment.
• Grow in a performance-driven workplace with ongoing learning and career development opportunities.
• Receive a competitive salary and company benefits.
• Work-from-home flexibility may be available depending on the team’s business needs and work nature.

Additional Information

Binance is an equal opportunity employer and values a diverse workforce as a core part of its success.

By submitting an application, you confirm that you have read and agree to the Candidate Privacy Notice.

Artificial intelligence tools may be used during parts of the hiring process, such as application review, resume analysis, or response assessment. These tools support the recruitment team but do not replace human judgment, and final hiring decisions are made by people. You may contact the company for more information about how your data is processed.