Data Scientist, LLM & AI Agent Engineer (Applied AI)

Overview

The company in question is a prominent player in the global blockchain landscape, known for its vast cryptocurrency exchange. Trusted by millions across various nations, the organization is renowned for its strong security measures, transparent fund management, swift trading engine, substantial liquidity, and diverse array of digital asset products. Their offerings encompass various aspects like trading, finance, education, research, payments, institutional services, Web3 functionalities, and more. Focused on leveraging digital assets and blockchain technology, the aim is to create an inclusive financial ecosystem that promotes financial freedom and enhances financial accessibility globally.

Key Role

As a pivotal member of the AI Risk team, you will play a crucial role in crafting specialized AI Agents that play a direct role in enhancing financial security. Your responsibilities will include employing LLMs (Large Language Models) to analyze unstructured data, automate intricate investigation workflows, and maximize the efficiency of alert systems.

Responsibilities

• Design autonomous agents akin to "Sherlock" for probing suspicious fraudulent activities
• Construct rationale-equipped agents for assessing the legitimacy of withdrawals by analyzing user behavior, device data, and transaction history
• Develop agents for detecting market manipulation by examining social sentiment data alongside order book trends
• Identify and combat schemes like pump-and-dump and wash trading
• Create triage systems based on LLMs to manage substantial volumes of risk alerts
• Utilize agents for pre-filtering alerts based on alert context and historical data to reduce false positives
• Automate the decision process of closing or escalating alerts to human analysts
• Employ LLMs to extract data from messy, unstructured sources such as chat logs, support tickets, etc.
• Translate vague risk signals into structured formats for downstream risk models or rule engines
• Build retrieval systems to enable agents to reference past fraud cases and recognize recurring patterns
• Develop tools for analysts to draft reports and summarize complex cases efficiently
• Facilitate human-in-the-loop workflows for enhanced collaboration and decision-making processes

Requirements

• Confirmed experience in applying LLMs like GPT, Claude, or similar models over at least 2 years
• Proficiency in agent frameworks like LangGraph, LangChain, or CrewAI
• Demonstrated skill in building retrieval systems for domain information using RAG
• Background in crafting evaluation sets to measure agent performance effectively
• Strong command over Python for developing production-grade, maintainable code
• Sound expertise in data engineering, including SQL and real-time data pipeline experience (e.g., Kafka, Spark)
• Bonus points for familiarity with Trust & Safety, anti-fraud, or financial risk domains, crypto knowledge, anomaly detection concepts, and financial systems understanding
• Educational background akin to a Master’s degree with hands-on expertise in Data Science, AI Engineering, or Applied ML
• Experience working with large datasets and real-time systems
• Skilled at driving execution in fast-paced, adaptive environments with a penchant for AI related technologies
• Effective English communication skills for technical documentation and agent interaction

Advantages of Joining

• Innovate within a world-class blockchain ecosystem
• Engage with a diverse, talented team in a global, user-focused environment
• Enjoy autonomy on challenging projects in an innovative setting
• Grow your career through results-driven initiatives and continuous learning opportunities
• Attractive package including competitive salary and benefits
• Flexible work arrangements to cater to different business team needs

The company upholds equal employment opportunities, recognizing the value of a diverse workforce in its success. If you apply for this role, you acknowledge having read and accepted the Candidate Privacy Notice. AI tools might be utilized during the recruitment process but final hiring decisions rest with human judgment. For further details on data processing, please feel free to reach out for more information.