尋找工作機遇

Nan Fung Group is building AI systems that change how our property, life sciences and financial investments businesses operate. Holonic is the group's AI team, sitting under the CEO Office and forward deployed to different parts of the business.

As a Forward Deployed Engineer you will sit with business teams, find the highest-value problems, and ship working AI solutions into production — fast. You own the full loop: discovery, prototype, deployment, adoption, iteration.

This is a hands-on builder role in a small, high-autonomy team. You will move quickly, work with ambiguity, and be judged on what actually gets used.

• Embed with business units across the group to identify workflows where LLMs and agents can deliver measurable impact
• Design and build production-grade AI systems: multi-agent workflows, RAG pipelines, tool/MCP integrations, evaluation and monitoring
• Take solutions from POC to deployment with real users, then iterate based on adoption and feedback
• Integrate with enterprise systems (ERP, CRM, data platforms) and legacy infrastructure
• Set engineering standards for reliability, guardrails and human-in-the-loop controls
• Translate between business stakeholders and technical implementation

What we're looking for

• 3+ years building and shipping AI/ML systems, with at least one LLM-based system running in production
• Strong Python; hands-on with agent frameworks (LangGraph, AutoGen or equivalent), RAG, vector databases and LLM APIs
• Experience deploying on cloud (AWS/Azure/GCP) with proper CI/CD, testing and monitoring
• Evidence of end-to-end ownership — you've gathered requirements, built, deployed and supported something real
• Comfortable in a fast-moving, loosely structured environment; you initiate rather than wait for direction
• Clear communicator with non-technical stakeholders; Cantonese and English

Nice to have

• Financial services, real estate or life sciences domain exposure
• Fine-tuning / evaluation experience (LoRA, custom evals, guardrails)
• Open-source contributions, side projects or founding experience
• Full-stack skills (React/Next.js, FastAPI) to ship complete tools

We offer

• Breadth of impact — unlike an FDE at a single-product company, you'll work across financial investments, life sciences, legal, finance, etc. often in the same month
• Autonomy — small team, direct access to business leaders, minimal bureaucracy between idea and implementation
• Visible transformation — you'll see the direct impact of what you build on how hundreds of colleagues work every day
• Learning velocity — exposure to the full stack of enterprise AI challenges: data integration, security, governance, change management, user adoption