About Aphelion
Where AI meets
considered practice.
We're a Hong Kong-based team that believes AI should serve people — not the other way around. Our work is grounded in craft, transparency, and a genuine respect for your context.
← Back to HomeOur Story
Built from a question, not a trend.
Aphelion emerged from a simple observation: most Hong Kong businesses were hearing a great deal about AI, but finding it difficult to translate that into practical action. The tools existed. The enthusiasm was there. What was missing was considered guidance.
We established Aphelion in Wan Chai to fill that gap — not as a technology vendor, but as a thoughtful partner. Our founding team came from analytics, enterprise software, and data science backgrounds, with shared experience working across Hong Kong's financial, trade, and professional services sectors.
From the beginning, our approach has centered on specificity. We don't offer generalized AI consulting. We build and assess AI systems for well-defined problems — the kind that your team encounters week after week. That focus shapes every project we take on.
Over time, we've come to see responsible AI not as a checkbox but as a design principle. It shapes how we scope work, how we handle data, and how we communicate our findings. We think that's what sustainable AI adoption actually looks like.
Years of Practice
Building and evaluating AI systems for Hong Kong businesses since 2020.
Client Engagements
Across finance, logistics, professional services, and retail sectors in the region.
Focused Services
Report automation, customer segmentation, and responsible AI review — each refined over many engagements.
The Team
The people behind the work.
Clara Leung
Co-Founder & AI Lead
Clara brings twelve years of applied machine learning experience, with a focus on natural language systems and enterprise data pipelines. She leads technical scoping and delivery on all client projects.
Marcus Tse
Co-Founder & Client Director
Marcus focuses on client relationships and project strategy. His background spans management consulting and analytics at firms serving the Asia-Pacific region, and he ensures every engagement is scoped with clear outcomes in mind.
Sunita Pillai
Responsible AI Practitioner
Sunita leads our responsible AI assessment work. She has a research background in algorithmic fairness and has contributed to published frameworks on AI accountability adopted in the financial sector.
How We Work
Standards we hold ourselves to.
Our methods are shaped by the belief that good AI work is careful work. These principles inform every project we undertake.
Data Confidentiality
All client data is handled under strict confidentiality protocols. We use encrypted transfers, access controls, and clear retention limits on every engagement.
Honest Scoping
We define project boundaries carefully and communicate plainly about what AI can and cannot do in your specific context. Clarity before commitment is a firm practice.
Fairness Standards
We examine AI outputs for bias and inequity as a matter of routine — not only in dedicated assessments, but as part of our general project quality reviews.
PDPO Alignment
Our data handling practices are designed to align with Hong Kong's Personal Data (Privacy) Ordinance. We can discuss specific compliance considerations at any stage of engagement.
Plain Communication
Technical findings are written to be understood by decision-makers, not just data teams. We avoid jargon where clarity matters, and explain our reasoning openly.
Iterative Delivery
We deliver incrementally and gather feedback throughout — so you stay informed and can shape the direction of the work as it develops, not only at the end.
AI expertise grounded in the Hong Kong context.
Working in Hong Kong's business environment shapes how we approach AI. The city's density of financial firms, trading operations, professional services, and international logistics creates a particular set of data challenges — fast-moving, compliance-aware, and often multilingual. Our services are designed with that context in mind.
Automated reporting, for instance, often intersects with regulatory filing requirements. Customer segmentation is most valuable when it reflects the behavioral patterns of Hong Kong's diverse consumer base. Responsible AI assessment gains particular weight when systems are used in credit decisions, recruitment, or client-facing automation.
We bring specific knowledge of these contexts to every engagement — alongside technical depth in machine learning, natural language processing, and analytical modeling. The combination allows us to offer advice that is both technically sound and practically relevant to your operating environment.
We'd welcome a conversation.
If you're curious about how AI might serve your team, we're glad to listen first and advise second.
Get in Touch