Head of Detection Real-Time Intelligence & Defense Systems

Head of Detection (Real-Time Intelligence & Defense Systems)

Job Description

Role Summary

We are seeking a Head of Detection to design and lead our real-time intelligence layer, responsible for identifying critical risks, anomalies, and opportunities across large-scale, fast-moving systems. This role leverages data, systems thinking, and AI to detect meaningful signals from vast, noisy, and seemingly unrelated data sources—enabling rapid downstream decision-making and automated action. You will be a core architect of our Defense Flywheel: Data → Signal → Decision → Action → Learning.

Key Responsibilities

\- Build a Unified Detection System:

\- Design detection frameworks across client behavior, system anomalies, human/operator anomalies, product & PnL irregularities, and cross-domain patterns.

\- Integrate multi-source data into a unified detection layer, including trading/activity logs, system metrics, user behavior, and financial outcomes.

\- Extract Signal from Noise (Core Mission):

\- Develop systems to identify non-obvious patterns across datasets.

\- Detect early weak signals and correlate multi-dimensional anomalies into actionable insights.

\- Build signal scoring frameworks to ensure output is actionable, high-confidence, and decision-ready.

\- Real-Time Detection Architecture:

\- Design and deploy low-latency detection pipelines.

\- Implement event-driven processing, streaming data systems, and real-time alerting frameworks.

\- Ensure high coverage, high reliability, and minimal detection delay (seconds-level).

\- AI & Model Integration:

\- Lead development of anomaly detection models, behavioral clustering, and pattern recognition systems.

\- Develop hybrid rule + ML detection frameworks.

\- Apply AI to reduce noise, improve precision, and discover hidden relationships.

\- Continuous Learning & Feedback Loop:

\- Build self-improving detection systems: incident → root cause → model refinement.

\- Own incident replay systems, pattern libraries, and model retraining pipelines.

\- Cross-Functional Signal Integration:

\- Partner with data engineering, infrastructure/system teams, risk/operations/trading.

\- Ensure detection logic reflects real-world system behavior.

\- Build & Lead Detection Team:

\- Hire and lead detection engineers, applied data scientists, and behavioral analysts.

\- Shift team mindset from “Monitoring & reporting” to “Real-time signal engineering”.

Required Skill Sets

\- 8–15+ years in real-time data systems, fraud detection/risk analytics, large-scale monitoring, AI/ML in production, distributed systems/platform engineering.

\- Experience with real-time anomaly detection platforms, monitoring systems at scale, high data volume, high noise, and high cost of delayed detection.

\- Systems Thinking: Ability to understand complex systems, cross-domain dependencies, and connect unrelated signals.

\- Data & Real-Time Processing: Experience with streaming systems (Kafka, Flink, Spark Streaming), event-driven architectures, and large-scale pipelines.

\- Applied AI / Detection Models: Expertise in anomaly detection, pattern recognition, behavioral analytics, and real-time deployment.

\- Signal Engineering: Ability to filter noise, design scoring, define dynamic thresholds, and prioritize signals.

\- Problem Decomposition: Skill in breaking down complex problems into structured detection logic and operating with incomplete information.

\- Fluency in both English and Chinese (Mandarin) is required for effective cross-regional communication and collaboration.