AI Research Engineer

XBOW is building the future of offensive security by creating an AI-powered platform that helps organizations discover and exploit vulnerabilities quickly. The AI Research Engineer will design and implement systems that coordinate large language models with real-world tasks, collaborating with various teams to create effective prompting strategies and production-ready systems.

Responsibilities:

• Build LLM-powered software that actually works, by designing prompt flows and orchestrations that ensures great performance with no false positives
• Architect and build an AI-powered software stack that is production-grade, testable and maintainable
• Design and build experiments and evaluation frameworks for performance testing of the system at scale. Conduct data analysis to draw conclusions
• Collaborate with the rest of the AI team, with security experts, and both frontend and backend developers to create end-to-end systems that work and customers love
• Own projects end-to-end: from basic ideation and experimentation to deployment and production monitoring
• Continuously conduct research on how to harness the advancements in LLMs to make our system better and faster

Requirements:

• Strong experience with building software around LLMs: prompting, agentic orchestration, fault-tolerance, and integration of LLM parts with hard-coded logic
• Strong software engineering skills: architecting and building production-grade software that runs reliably and can be maintained
• Experience with TypeScript or proven ability to learn a new programming language quickly
• Strong skills in structured and independently-driven problem-solving. Able to work with incomplete information and rapidly testing hypotheses
• Comfortable with an energetic environment that mixes the fast-paced agile prioritisation of a startup with the curiosity mentality of a research lab
• Eager to own projects and jump into the deep end, learning as you go. Curious, adaptable and collaborative
• MSc or equivalent or higher in computer science, math, physics or machine learning
• Prior experience with the security sector, especially involving offensive security and/or applying LLMs in a security context
• Contributions to open-source software, particularly within applications of or frameworks for LLMs
• Strong theoretical background, e.g. a Ph.D in machine learning, computer science or math
• Startup experience and comfort with ambiguity