Key Responsibilities

• Agentic Architecture: Design and implement multi-agent systems (planning, tool use, routing, memory, orchestration, guardrails).

• R&D and Fine-Tuning: Evaluate emerging models, fine-tune open-source LLMs for specific financial tasks, and optimize prompt + system architecture for highly reliable outputs.

• Evaluation & Reliability: Build robust evaluation pipelines, regression tests, and monitoring to ensure consistent performance in real-world conditions.

• AI Infrastructure: Deploy and scale LLM/agent workflows (serving, latency/cost optimization, observability, incident response as needed).

• Collaboration: Work closely with product/engineering stakeholders to translate financial workflows into measurable AI capabilities.

• Experience in financial services (trading, risk, compliance, research, reporting, KYC/AML, etc.).

Key Requirements

• Bachelor’s degree in computer science, Information Systems, or a related field.

• 5+ years of software engineering experience with deep, recent focus on LLMs, agentic workflows, and AI infrastructure.

• Strong hands-on experience with AI evaluation frameworks and designing testable, measurable AI systems.

• Mindset: researcher at heart, hacker in execution—curious, experimental, but able to ship production-grade systems.

• Strong logical thinking, proactive problem-solving, and the ability to work in an agile environment.

• Proficiency in English and Cantonese.