Detection Expert

Role Overview

Hytech is seeking a seasoned Detection Product Expert charged with the design, enhancement, and ownership of our real-time detection capabilities. This senior individual contributor role entails being the authoritative expert and steward of detection logic, working at the convergence of data engineering, AI/ML, and risk operations. Although this role does not include direct team management responsibilities, it plays a critical part in preserving platform integrity and safety.

Key Responsibilities

• Lead the comprehensive ownership of the detection product—including setting detection logic, signal scoring methods, and dynamic thresholding across user behaviors, anomalies in the system, and financial inconsistencies.
• Derive actionable insights from vast, noisy, multi-faceted data sources such as trading logs, user interactions, system data, and financial outcomes.
• Architect and fine-tune real-time detection pipelines that meet stringent low-latency requirements, ensuring sub-second to seconds-level responsiveness for critical events.
• Create and enhance anomaly detection algorithms, behavioral clustering, and pattern recognition frameworks by integrating hybrid methodologies combining rules and machine learning.
• Establish and maintain continuous feedback cycles that move from incident analysis to root cause discovery, model tuning, and retraining.
• Translate intricate detection needs into clear technical specifications used by engineering teams, effectively bridging detection logic and operational system implementation.
• Collaborate cross-functionally with data engineering, risk management, trading desks, and infrastructure teams to guarantee comprehensive system coverage and alignment.
• Focus on minimizing false positive alerts whilst preserving strong detection accuracy and coverage across system and user activities.

Qualifications & Skills

• 5 to 10 years working with real-time detection systems, fraud or risk analytics, trade surveillance, or applied AI/ML in production settings.
• Proven hands-on experience developing and owning anomaly or detection systems end-to-end, beyond simply contributing components.
• Expertise in machine learning and AI techniques specifically for anomaly detection, behavioral analytics, and pattern recognition, including real-time deployment.
• Advanced skills in Python and SQL, alongside experience with event-driven or streaming data technologies such as Kafka, Flink, KDB/q, or comparable platforms.
• Strong systems thinking abilities to connect disparate signals, analyze complex dependencies, and break down ambiguous challenges into organized detection strategies.
• Fluency in English and Mandarin Chinese, enabling effective collaboration across regions.

Preferred Experience

• Background in finance, cryptocurrency exchanges, fintech solutions, or managing risk on large-scale internet platforms.
• Knowledge in graph-based detection systems, including knowledge graphs and fraud network analytics.
• Demonstrated success in reducing false positives, improving detection speed, or lowering fraud losses.
• Experience with market manipulation identification, trade surveillance, or risk monitoring at the exchange level.