Senior Engineer, AI Agent Security Research at OKX

Singapore, , Singapore -

Full Time

Start Date

Immediate

Expiry Date

28 May, 26

Salary

0.0

Posted On

27 Feb, 26

Experience

2 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Skills

Python, Go, Java, LLM Agents, AI Security, Prompt Injection, Jailbreaking, LangChain, LlamaIndex, AutoGen, CrewAI, LangGraph, Docker, Kubernetes, Microservices, RAG

Industry

Financial Services

Description

OKX will be prioritising applicants who have a current right to work in Singapore, and do not require OKX's sponsorship of a visa. Who We Are At OKX, we believe that the future will be reshaped by Crypto, ultimately contributing to every individual's freedom. OKX began as a crypto exchange giving millions of people access to crypto trading and over time becoming among the largest platforms in the world. In recent years, we have developed one of the most connected Web3 wallets used by millions to access decentralized crypto applications (dApps). OKX is a trusted brand by hundreds of large institutions seeking access to crypto markets on a reliable platform that seamlessly connects with global banking and payments. In the last year, OKX has expanded into new markets including Australia, Brazil, Netherlands, Singapore and Turkey, with plans to launch in the US, Belgium and the UAE. We are deeply committed to shaping a fairer, more transparent and accessible society through blockchain technology. This is why we publish proof of reserves monthly, and continue to ship new innovative security features. What You’ll Be Doing AI-Driven Code Security Detection Engine Design and implement a multi-agent collaborative code auditing system covering vulnerability detection, malicious code identification, and sensitive information leakage scenarios; lead the role decomposition of Planners/Executors/Critics, tool invocation chains, and cross-agent state synchronization mechanism design. Integrate RAG, Chain-of-Thought, Reflection, and other technologies into security audit agents. Continuously optimize detection accuracy and recall rates while establishing a quantifiable evaluation and iteration framework. Deeply integrate with DevSecOps workflows. Develop plugins for mainstream pipelines like GitLab CI/CD, Tekton, and Jenkins to achieve “audit-on-commit.” AI System Security Protection and Threat Response Responsible for constructing a security protection framework for large language model applications, covering three dimensions: input layer (prompt injection, jailbreak detection), output layer (sensitive information leakage, compliance auditing), and runtime (tool invocation sandboxing, anomaly behavior circuit breaking). Develop Agent workflows for automated alert classification, contextual correlation, and false positive filtering. Integrate RAG-driven threat intelligence retrieval to generate automated analysis conclusions, supporting SOAR platform integration. Design human-machine collaboration intervention mechanisms and Agent behavior audit systems to ensure observability, traceability, and intervenability of Agent actions in production environments, adhering to industry standards like the OWASP Top 10 Risks for LLMs. Engineering Development and Platform Services Construct a highly available, scalable Agent service architecture supporting large-scale concurrent scanning task scheduling and fault tolerance. Oversee standardized API output for detection capabilities, building closed-loop systems for rule management, result visualization, and false positive feedback. What We Look For In You Development Experience: 3+ years of backend development experience, proficient in at least one of Python/Go/Java, with a solid engineering foundation. Agent Implementation & Security: Hands-on experience deploying LLM Agents (not just demos), capable of detailing engineering challenges such as Agent architecture design, hallucination handling, and tool invocation fault tolerance; Hands-on experience with AI security, understanding risks like prompt injection, jailbreaking, malicious agent injection, and tool misuse, with implementable defense strategies. Framework Proficiency: Familiarity with at least one agent framework (LangChain, LlamaIndex, AutoGen, CrewAI, or LangGraph), with production project experience. Engineering Capabilities: Proficient in Docker and Kubernetes, with expertise in microservices architecture design and deployment. Nice to Haves Security Tool Experience: Experience with SAST/SCA tools, or deep usage of code auditing tools like CodeQL, Semgrep, or SonarQube. Model Fine-Tuning: Experience with LLM fine-tuning (SFT, LoRA), or familiarity with local deployment and optimization of models like Llama 3, Qwen, or DeepSeek. Bonus points for security-domain fine-tuning experience, such as training and evaluating security detection models for malicious prompt detection, unauthorized access identification, or harmful content filtering. Open-Source Contributions: High-quality open-source projects related to agents on GitHub, or pull requests submitted to mainstream LLM frameworks. Security Competitions: Awards from CTF competitions, or a track record of submitting CVE/CNVD vulnerabilities. Perks & Benefits Competitive total compensation package L&D programs and Education subsidy for employees' growth and development Various team building programs and company events Wellness and meal allowances Comprehensive healthcare schemes for employees and dependants More that we love to tell you along the process! #LI-ML1 #LI-ONSITE Notice: All official OKX vacancies are published on this website. While roles may appear on selected third-party platforms from time to time, information on other sites may be inaccurate or outdated. If in doubt, please apply directly through our official careers website. Information collected and processed as part of the recruitment process of any job application you choose to submit is subject to OKX's Candidate Privacy Notice.

Responsibilities

The role involves designing and implementing a multi-agent collaborative code auditing system for vulnerability detection and malicious code identification, integrating advanced AI techniques like RAG and Chain-of-Thought. Additionally, the engineer will construct a security protection framework for LLM applications covering input, output, and runtime dimensions, and develop Agent workflows for automated threat response.