Lead AI Engineer - AI Assistant Kai

About the role

Airwallex, the AI-native financial operating system used by more than 675,000 businesses, is hiring a Lead AI Engineer in its Singapore office. You will lead the technical design, execution and people leadership of Kai, Airwallex's AI assistant for complex financial work, turning rapidly evolving LLM and agent capabilities into reliable, observable and safe production systems.

Key responsibilities

• Set the technical direction for agent orchestration, context and retrieval, tool integrations, identity and access control, observability, evaluation and responsible action-taking, balancing quality, latency, reliability, cost, security and time to market.
• Work with Product, Design, Data, Security, Infrastructure and domain teams to deliver trusted AI products and reusable platforms that can scale across Airwallex.
• Manage, coach and develop AI, backend and frontend engineers, set clear expectations and raise the engineering bar through hiring, feedback and technical mentorship.
• Build intelligent financial products and autonomous agents that understand business context, plan and reason across multi-step workflows and execute actions within clearly defined controls.

Required profile

• 8+ years of software engineering experience, including 2+ years leading, managing or formally mentoring engineers.
• Track record of building strong teams and delivering complex products end to end.
• Deep experience designing and operating scalable agentic architecture, backend systems or distributed systems.
• Comfortable owning the roadmap yourself, working well with cross-functional teams and communicating technical decisions and risks clearly.

Required skills

• Understanding of retrieval, model integration, agent workflows, evaluation, observability and changing model behaviour.
• Secure software design, including authentication, authorisation, data protection, auditability and abuse prevention.
• Ability to define and improve SLOs for critical AI services and to build versioned evaluation practices for task success, grounding, tool use and safety.
• Experience defining launch criteria, staged rollouts, rollback mechanisms and monitoring that detects regressions.